Discourse of persuasion: a preliminary study of the use of metadiscourse in policy documents

1 Introduction

Since the return of its sovereignty to China in 1997, Hong Kong has witnessed changes in various aspects such as the healthcare system and the political ecology both inside and outside of the administration. These changes, to a certain extent, can be attributed to the sovereignty change. The public healthcare system of Hong Kong has been described as unsustainable (The Harvard Team 1999) because of the provision of high-quality services to patients heavily subsidized by the HKSAR government. The problem has been aggravated by the influx of healthcare service users from mainland China in recent years. The HKSAR government has thus been promoting a healthcare reform which aims to adjust the imbalance between public and private healthcare sectors, and to ensure sufficient funding for the sustainable operation of the public healthcare system by getting people to pay for the public healthcare services they receive. The political ecology both inside and outside of the administration has become more diversified in that new political parties with distinct ideologies have emerged, leading to frequent confrontations between the government and the parties, or between the parties themselves.

The public in general has also become more politically sensitive since the sovereignty change. An increasing number of people now show a greater interest in politics and a greater willingness to participate in political activities ranging from peaceful assemblies and rallies to radical demonstrations and protests to show their anger and dissatisfaction with the administration, to fight for their own rights, or to make their voices heard. The increasingly politically aware and politically sensitive population has made the government’s proposed reform to the political structure in 2012, and its proposed way of filling a vacancy in the Legislative Council following the resignation of a legislator in 2011, the subject of heated debate among the public. Both proposals have drawn severe criticisms across the territory. With a population with increasing political interest and awareness, together with other factors reported in a recent study (Ho 2011), such as the Hong Kong people’s negative perception of their government and the victimization philosophy (Carroll and Buchholtz 2009), it is reasonable to assume that the HKSAR government will need to employ means of persuasion in its policy reform consultative documents to convince the people of Hong Kong about the need for the proposed reforms (Ho 2011). That is, the policy document genre should be persuasive and argumentative in nature (Saarinen 2008).

The present study aims to explore how the HKSAR government attempted to persuade Hong Kong people to accept two proposed education policy reforms: the fine-tuning of the medium of instruction (MOI) policy in secondary schools in 2009, and the incorporation of national education (NE) in the curriculum of local secondary and primary schools in 2012. The MOI reform aims to allow secondary schools to decide which language, Chinese or English, they should use as the medium of instruction; and the NE reform aims to make national (about China) education a compulsory subject in local schools. The HKSAR government would need to persuade the public through its consultative documents that first, the current education system in relation to the medium of instruction policy and the absence of a formal, mandatory national education curriculum is problematic; and second, that the proposed reforms can address such problems and issues (Saarinen 2008).

The present study has two research questions:

1. How does the metadiscourse use pattern in the policy reform genre compare to that in other genres?

2. How does the HKSAR government attempt to persuade the readers of the policy reform documents to accept the proposed reforms with metadiscourse?

These two reforms are chosen since both concern education, and thus the discourse of the consultative documents should be constructed by the members of the same discourse community – the Administrative Officers of the Education Bureau of the HKSAR government. The effect of the differences in norms, practices, and culture of different discourse communities (i. e. government bureaus) can then be minimized.

2 Language in policy reforms

While policy discourse has been approached from a positivistic perspective which sees a direct relationship between the nature of the world and the conventional knowledge of such nature as portrayed through language in the past (e. g. Fischer and Forester 1993; Fox and Miller 1995), more recent studies tend to approach it from a social constructionist perspective which sees the nature of the world as constructed with language. As such, the problems and issues of the existing policies and the solutions to such problems and issues presented as the reform initiatives, should be seen as constructed by the government officials with discourse. Such problem–solution discursive construction could serve to persuade people to accept the proposed reform initiatives (Ho 2011).

Discursive construction of this type has been demonstrated in previous studies. Hastings (1998) discussed the way the Scottish government framed the population growth in urban Scotland as a problem in that the growth was inexorable and would probably lead to imminent disaster. The problem was constructed with the following clauses: “Nowhere was this more apparent than in Glasgow where the population rose from 500,000 in 1871 to 750,000 in 1891 and topped 1,000,000 in 1914. The size and density of inner city populations rose to intolerable levels” (The Scottish Office 1988: 6; emphasis mine). Hastings (1998: 201) argued that the Scottish government could have described the growth in an objective, factual manner as in “[t]his was particularly the case in Glasgow where the population doubled between 1871 and 1914.” Osgood (2009) focused on the policy concerning the childcare workforce reform in the United Kingdom and found that the UK government portrayed the childcare workers as deficient as in “failing to meet the needs of many children and parents as society has changed”, and “under-qualified” (DfES 1997: 10–11). The government then presented the solution as in “[g]overnment action is needed” (DfES 1997: 7). Rogers (2012) described how politicians and the State of Missouri Department of Education officials constructed a problem for the public schools in St. Louis in order for the state to take over those schools. The mayor of St. Louis gave this remark on a public occasion: “The public schools are in a state of crisis” (Slay 2006) constructing a problem for the public schools; and the Education Commissioner announced that the School Advisory Committee would be responsible for making “recommendations regarding the future direction of the school district” and would “recommend any changes in legislation that might be needed to implement its recommendations” as the solution to the problem constructed (Department of Elementary and Secondary Education 2009).

Some other studies discussed how national governments attempted to persuade people to accept reforms through other discursive means. Fairclough (2000) analyzed the lexicogrammar of the welfare system reform documents of the UK government and found that the government achieved persuasion by, among other methods, constructing differing identities for itself and the welfare recipients. It depicted itself as an entity with authority and power by making itself the agent of clauses with “actional process” (Fairclough 2000: 177), and depicted the welfare recipients as trivial figures lacking initiative by making them the beneficiary of clauses. Flowerdew (2004) showed how the HKSAR government attempted to persuade not only the people of Hong Kong, but also foreigners around the world that the city was a World City. It was found that, among other lexicogrammatical and semiotic resources, the government frequently resorted to interpersonal resources which included the pronouns I, we, and our; questions which “immediately draw readers into the text” (Flowerdew 2004: 589); and the semantic realization of the “repeated emphasis on participation” (Flowerdew 2004: 589). In their study of the curriculum reform in Hong Kong, Morris and Chan (1997) found that the then Education Department attempted to ideologically persuade and motivate the teachers to accept and implement the reform by empowering them as in “the class teacher is the key person […] to instill in his pupils the proper attitudes and right sets of values” (Education Department 1981: 9). In another study concerning education policy reform, Saarinen (2008: 345) discussed the use of informative presuppositions by OECD (Organization for Economic Cooperation and Development) and European Union in persuading people to accept higher education policy reforms.

While the strategies reported above have been shown to be persuasive, that is, people were convinced that the need for the proposed reforms was real, what is yet to be explored systematically in policy discourse is the use of the three means of persuasion which have “characterized persuasive discourse since the time of ancient Greece” (Hyland 2005: 63), namely ethos, pathos, and logos, all of which can be realized through the use of metadiscourse. The construct of metadiscourse and these means of persuasion will be discussed in the next section.

3 Metadiscourse, persuasion, and genre

A number of definitions of metadiscourse have been put forward by scholars at different periods. Both Williams’s (1981) and Vande Kopple’s (2002) definitions see discourse as comprising two distinct components – one primary and the other secondary, referred to respectively as the subject matter/propositional content (primary), and the language we use to express our stance and attitude toward that subject matter/non-propositional, i. e. metadiscourse, content (secondary). However, this two-level distinction has been challenged as omission of metadiscourse would somehow change the meaning of a text (Hyland and Tse 2004). Crismore and Farnsworth’s (1990) and Hyland’s (2005) definitions shift away from a primary-secondary distinction and see metadiscourse as a resource for the writer to intrude into the discourse and to interact with the reader. Both Crismore and Farnsworth (1990) and Hyland (2005) agree that people make deliberate choices while using language in interacting with others. This paper will use Hyland’s (2005) definition of metadiscourse as it emphasizes the interactional – writer and reader are constantly interacting with each other while a text is being produced and consumed: “Metadiscourse is the cover term for the self-reflective expressions used to negotiate interactional meanings in a text, assisting the writer (or speaker) to express a viewpoint and engage with readers as members of a particular community” (Hyland 2005: 37). Hyland’s (2005) interpersonal model of metadiscourse is shown in Table 1.

The use of metadiscourse has often been regarded as the writer’s attempt to achieve persuasiveness through making rational (logos), credible (ethos), and affective (pathos) appeals to the readers. It has been shown that ethos (i. e. appeals to credibility and character) can be achieved by the use of hedges and boosters; logos (i. e. using reason in persuasion) can be achieved by linking the arguments logically with interactive metadiscourse; and pathos (i. e. affective appeals) could be achieved by “looking at the text from the readers’ perspective, addressing their situation, empathizing with their values and goals, and directly inviting them to respond” (Hyland 2005: 81) with interactional metadiscourse.

Crismore and Farnsworth (1989) demonstrated how persuasiveness in science texts could be achieved with attitude markers and the balanced use of hedges and boosters. Dafouz-Milne (2008) compared metadiscourse use in two newspapers, one Spanish and one English, and concluded that both interactive and interactional metadiscourse were necessary if one was to succeed in convincing others. Virtanen (2005) also explored how persuasion was achieved through ethos in newspapers – editorials. She found that editors used vagueness and assertiveness when making references to polls and surveys. Vagueness was expressed through the use of hedges (e. g. suggest, seem to, may) when the polls and surveys were referred to; and assertiveness was expressed through the use of boosters (e. g. show) in the poll clause. Hyland (1998a) examined business discourse – CEOs’ letters in published annual reports of listed companies – and found frequent use of metadiscourse in achieving persuasiveness by creating rational, credible, and affective appeals.

The majority of literature on metadiscourse, however, focuses primarily on business genres and various academic genres. Other genres have comparatively been less attended to by the research community. The use of metadiscourse in realizing means of persuasion in policy documents, to the best knowledge of the author, has not been explored. The present study represents a modest attempt to address and bridge this gap by analyzing the discourse of the policy documents of two recent education reforms in Hong Kong – the fine-tuning of the MOI (medium of instruction) policy in 2009 and the incorporation of NE (national education) in the curriculum of local schools in 2012. The focus will be on the use of metadiscourse in realizing the three means of persuasion in the two sets of policy documents.

I find it useful to include a brief discussion of genre here before proceeding to the next section which describes the methods in analyzing the policy reform genre. According to Bhatia (2004), genres make it possible for professionals to perform their everyday tasks which include getting one’s colleague to perform an act with the request e-mail genre (Ho 2014), persuading potential customers to purchase a product or a service with the promotional genre (Vergaro 2004), and to disseminate research findings and establish the author’s academic credentials with the research article genre (Lin and Evans 2012). According to Bhatia (2004: 23; emphasis mine), “[g]enre essentially refers to language use in a conventionalized communicative setting in order to give expression to a specific set of communicative goals of a disciplinary or social institution, which give rise to stable structural forms by imposing constraints on the use of lexico-grammatical as well as discoursal resources.” Drawing upon genre theory (Bhatia 2004), researchers have examined the “stable structural forms,” the move structure, of various genres in the past few decades. The academic genre, business genre, and promotional genre are three of the most extensively researched genres. Apart from the move structure, “the use of lexicogrammatical as well as discoursal resources” has also attracted considerable research interest. In her study of the business report genre, Yeung (2007: 166) found that report writers created “an impression of professionalism and objectivity” with nominalization, expressions contributing to the formation of rational arguments and impersonality, evaluative language, and tones. Sadeghi and Samuel (2013) examined the tense and request strategies used in letters of appeal. Zhou (2012) investigated the hybrid genre advertorials and reported the use of evaluative resources – attitude, engagement, and graduation (Martin and White 2005) – in the genre. The present study will focus on this latter attribute of the policy reform genre – the lexicogrammatical and discoursal resources, and in particular, the use of metadiscourse by the HKSAR government in its attempt to persuade the people of Hong Kong to accept the two proposed education policy reforms.

4 Methodology

The data of the study comprised the documents concerning the two education policy reforms – the fine-tuning of the Medium of Instruction policy in 2009 and the incorporation of National Education in the curriculum of local schools in 2012 (please see the appendix). The website of the Education Bureau was searched comprehensively for the documents released by the HKSAR government concerning the two reforms. These documents varied in a number of aspects: some were written to be read and some were written to be spoken (speeches, and Legislative Council’s questions and replies); some were consultation documents and some were letters and memoranda; some were meant to be read by the general public and some were meant to be read by some particular groups (e. g., school principals, teachers, and Legislative Councilors). A total of twelve documents concerned the MOI policy and only four concerned the incorporation of NE policy. The difference was mainly a result of the fact that the HKSAR government had only published the other NE documents in Chinese and these documents, therefore, were not analyzed in the present study. Two corpora, one for each set of documents, were then created using Wordsmith 6.0. The total number of words of the MOI policy documents was 81,297, and that of the NE policy documents was 85,493.

With the Concord function of Wordsmith 6.0, the two corpora were searched to locate the potential metadiscourse markers suggested by Hyland (2005: 218–224). Each of these markers in context was then examined manually to determine whether it was metadiscursive or propositional for the fact that these markers can “refer to either relationships internal to the discourse or to events in the world” (Hyland 2005: 46). As metadiscourse only concerns those features which refer to text-internal relationship, the identified potential metadiscourse markers were scrutinized closely to decide whether they relate statements in the text to the readers, i. e. metadiscursive, or serve to relate statements in the text about the world, i. e. propositional. Extracts (1a) and (1b) below, taken from the data with added emphasis, will illustrate the distinction. Both contain the same potentially metadiscursive expression could.

The writer of (1a) was explaining the observed difference in students’ performance in science subjects and social studies on a public examination. S/he used could in the explanation concerning the science subjects, but used was mainly and thus sounded more certain in the explanation concerning social studies, suggesting strongly that s/he was deliberately withholding her/his commitment to the explanation in the former explanation. The modal could should therefore be metadiscursive.

The writer of (1b) was discussing the issue of discretionary places (DP) – places each secondary school principal can admit without following the Education Bureau’s central student allocation mechanism. The writer should be referring to an issue in the real world when using could in the first sentence, that is, most schools and parents found the DP percentage insufficient. Such referral can be verified when we look at the second sentence in which the writer suggested a possible solution – the writer should not have made the suggestion if the issue had not existed. Therefore, the modal could in (1b) was propositional.

I am aware that the distinction between metadiscursive and propositional is “fuzzy” (Khabbazi-Oskouei 2013: 93) and thus relying on one single list for identifying metadiscourse markers is not enough. I therefore did the identification by manually reading the text closely, analyzing the function of the words, phrases, and even clauses to determine if they were metadiscursive.

5 Findings and discussion

This section will report and interpret the observed metadiscourse use patterns at the macro level and micro level. At the macro level, reference will be made to the patterns observed in other seminal metadiscourse studies exploring different fields and genres. At the micro level, the use of metadiscourse in the policy documents in an attempt to persuade the readers will be discussed.

6 Conclusion

This paper has attempted a corpus-based analysis of the policy reform genre which, to the knowledge of the author, has not yet been approached from a metadiscursive perspective. The analysis shows that the government officials have made persuasive attempts by appealing to ethos, pathos, and logos. The use of metadiscourse was first examined at both the macro and micro levels. At the macro level, in comparison with the writers of the other genres (CEO letters, director reports, newspaper opinion columns, and research articles) in relation to the use of metadiscourse in effecting these three means of persuasion, those of the education policy documents genre exhibited a medium tendency in that they used both interactive and interactional metadiscourse more frequently than the businessmen and less so than the opinion columnists and researchers. When examined at the micro level, it is found that three interactional metadiscourse markers – hedges, engagement markers, and self-mentions – were used frequently in establishing credibility for the government (i. e. ethos) and achieving affective appeals (i. e. pathos). Similarly, three interactive metadiscourse markers – transition markers, endophoric markers, and code glosses – were used frequently in persuading with reasons (i. e. logos).

The present study has both theoretical and practical significance. Theoretically, it has extended the scope of analysis of metadiscourse as an analytical framework. Metadiscourse has mainly been drawn upon in the analysis of the business, academic, and journalistic discourse. The present study has used it in analyzing the policy discourse. Practically, the study has made the use of metadiscourse in effecting persuasive attempts in the two sets of education policy reform documents explicit on the one hand, and informed government officials involved in the construction of the policy document genre of some of its salient linguistic features on the other. The latter contribution should be of significance to the HKSAR government, considering first the increasingly politically aware and politically sensitive Hong Kong society, and thus, second, the need for the government to persuade the people of Hong Kong to accept the various policies that it will propose in the future. If the officials responsible for producing the genre can be made aware of the linguistic resources, metadiscourse in this case, that they can use to make their writing persuasive, the HKSAR government would stand a better chance to convince the people of Hong Kong to accept its proposed policies or policy reforms.

Despite its contribution in regard to the advancement of theory and understanding of professional practice, the study can be refined and extended in three ways. First, the differences in the use of metadiscourse between the two sets of education policy reform documents can be revealed and discussed in relation to the fate of these two reforms – the smooth implementation of the one concerning the fine-tuning of the medium of instruction and the complete withdrawal and eventual shelving of the one concerning the incorporation of the national education in secondary and primary school curriculum. Second, policy documents concerning areas other than education (e. g. healthcare, housing, welfare) should be analyzed. Third, upon completion of more empirical studies, it should then be possible for researchers to ascertain whether or not metadiscourse and its use in the context of public policy discourse is “integral to the contexts in which it occurs and is intimately linked to the norms and expectations of particular cultural and professional communities” (Hyland 1998b: 438), thereby enhancing our understanding of the public discourse genre.