Scholarly article on topic 'Metadiscourse in the introductions of PhD theses and research articles'

Metadiscourse in the introductions of PhD theses and research articles Academic research paper on "Languages and literature"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{"Genre analysis" / Rhetoric / Metadiscourse / "Doctoral thesis or dissertation" / "Research article" / Introduction}

Abstract of research paper on Languages and literature, author of scientific article — Tomoyuki Kawase

Abstract Previous studies have indicated that the introductions of PhD theses and research articles are similar in their rhetorical features. In contrast, it has been suggested that metadiscourse as a rhetorical device is constructed in a different manner in these texts. However, very few studies have sought to empirically validate this assumption. This paper investigates how research writers construct metadiscourse in the introductions of their PhD theses and subsequently published research articles. The analysis shows that the majority of the writers make greater use of metadiscourse in their article introductions. The most significant changes include greater use of phrases referring to previous research, less reference to other parts of the text, and still less use of phrases signalling authorial presence. Close examination reveals that these variations derive from genre-specific features, including that writers of PhD thesis introductions present previews of the subsequent chapters. This paper closes by arguing that the variations can also be ascribed to the nature of the PhD thesis as an educational genre and that of research articles as a professional genre in which writers need to survive severe competition to get their manuscripts published.

Academic research paper on topic "Metadiscourse in the introductions of PhD theses and research articles"

Contents lists available at ScienceDirect

Journal of English for Academic Purposes

journal homepage: www.elsevier.com/locate/jeap

Metadiscourse in the introductions of PhD theses and research articles

Tomoyuki Kawase*12

Faculty of Arts and Education, Deakin University, 221 Burwood Highway VIC. 3125, Australia

ARTICLE INFO

ABSTRACT

Article history:

Received 9 July 2014

Received in revised form 19 August 2015

Accepted 24 August 2015

Available online xxx

Keywords: Genre analysis Rhetoric Metadiscourse

Doctoral thesis or dissertation Research article Introduction

Previous studies have indicated that the introductions of PhD theses and research articles are similar in their rhetorical features. In contrast, it has been suggested that meta-discourse as a rhetorical device is constructed in a different manner in these texts. However, very few studies have sought to empirically validate this assumption. This paper investigates how research writers construct metadiscourse in the introductions of their PhD theses and subsequently published research articles. The analysis shows that the majority of the writers make greater use of metadiscourse in their article introductions. The most significant changes include greater use of phrases referring to previous research, less reference to other parts of the text, and still less use of phrases signalling authorial presence. Close examination reveals that these variations derive from genre-specific features, including that writers of PhD thesis introductions present previews of the subsequent chapters. This paper closes by arguing that the variations can also be ascribed to the nature of the PhD thesis as an educational genre and that of research articles as a professional genre in which writers need to survive severe competition to get their manuscripts published.

© 2015 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Previous studies of academic writing have often indicated that the introductions of PhD theses and research articles (RAs) are similar in their rhetorical features (e.g., Bunton, 1998, 2002; Feak, Swales, & Irwin, 2011; Paltridge & Starfield, 2007; Samraj, 2008; Swales, 2004). Swales (2004), for instance, wrote that the overall structure of the introduction of PhD theses "in broad outline is comparable to that of research articles" (p. 117). Samraj (2008) noted that "[s]tudies of introductions in research articles and PhD theses have made it clear that a crucial rhetorical function of introductions is to justify the study being reported" (p. 56). However, it has been suggested that writers of these introductions are dissimilar in terms of how they construct the rhetorical device of metadiscourse or "self-reflective linguistic expressions referring to the evolving texts, to the writer, and to the imagined reader of that text" (Hyland, 2004, p. 133; see also Adel, 2006; Crismore, 1989; Crismore, Markkanen, & Steffensen, 1993; Mauranen, 1993). Swales (1990) claimed that "the key differentiating aspect of

* Current address: 59 Tono Inuigaito, Joyo, Kyoto, Japan. E-mail address: tkawjpbon@gmail.com.

1 This is where the actual work was done; it is not the author's current affiliation.

2 Tomoyuki Kawase is a researcher for whom English is a foreign language. He completed his PhD studies at Deakin University in Australia. His thesis explores Japanese researchers' development of academic literacy in English. His research interests also include the roles of rhetoric in knowledge construction.

http://dx.doi.org/10.1016/joeap.2015.08.006

1475-1585/© 2015 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).

dissertation writing [from article writing] is a much greater use of metadiscourse" (P. 188). Bunton (1998) analysed 13 PhD theses and presented findings to explicitly support Swales' claim. Paltridge and Starfield (2007, p. 89) attributed thesis writers' greater use of metadiscourse to "much lengthier" text characteristic of the thesis genre. Swales (1990) drew the same conclusion by saying that "[m]etadiscourse goes with extensive textual territory" (p. 189). Studies examining more specific usage of metadiscourse in PhD theses and RAs also tend to indicate variations derived from genre-related factors. Koutsantoni (2006), for instance, argued that research writers change how to use metadiscourse items to control the strength of their claims according to their relationship with the intended audience of the genre (e.g., thesis examiners and journal reviewers) (cf. Shaw, 2000).

It should be noted that none of the conclusions drawn in the above studies are about the metadiscourse use in the introductions alone. Moreover, to my knowledge, no study has empirically sought to examine assumptions regarding how and why metadiscourse is constructed differently in the introductions of PhD theses and RAs based on the comparative analysis of both texts.

This paper explores how eight writers construct metadiscourse in the introductions of their PhD theses and RAs that they later produced based on the theses. By doing this, it examines the assumption that variations in the use of metadiscourse in these texts can be attributed to the nature of the genre. The findings will provide useful insights, especially for early career researchers aspiring to publish RAs based on their theses.

2. Material and methods

2.1. Definitions of metadiscourse

Previous studies have often classified metadiscourse in relation to the three communicative functions of language identified by Hallidayan systemic functional grammar (e.g., Halliday, 1994), as has been pointed out (e.g., Adel, 2006; Hyland, 2005). They are explained by Hyland (2005, p. 26) in the following terms:

• The Ideational function: the use of language to represent experience and ideas

• The Interpersonal function: the use of language to encode interaction, allowing us to engage with others, to take on roles

and to express and understand evaluations and feelings

• The Textual function: the use of language to organize the text itself, coherently relating what is said to the world and to

readers.

It appears that the majority of metadiscourse theorists (e.g., Crismore et al., 1993; Hyland, 2000; Hyland & Tse, 2004; Vande Kopple, 1985, 2002) have adopted the notion that metadiscourse does not serve an ideational function (i.e., to construct propositional content) but textual and interpersonal functions. In contrast, studies like Adel (2006) and Mauranen (1993) consider that metadiscourse contains (meta)textual items alone, questioning the notion that metadiscourse consists solely of non-propositional items.

We should note that separate analyses of metadiscourse items with single functions could lead to miss the possible interactions between them (cf. Crismore et al., 1993; Hyland & Tse, 2004; Vande Kopple, 2002). This point is suggested, for example, by Hyland as follows:

In other words, while metadiscourse theorists tend to see textual, interpersonal and propositional (ideational) elements of the texts as discrete and separable, Halliday reminds us that texts have to be seen more holistically. Discourse is a process in which writers are simultaneously creating propositional content, interpersonal engagement and the flow of text as they write. (Hyland, 2005, p. 27)

Nevertheless, previous studies suggesting thesis writers' greater use of metadiscourse in the introductions (e.g., Bunton, 1998; Paltridge & Starfield, 2007; Swales, 2004) tend to assume that metadiscourse contains only metatextual items. Swales (2004) wrote that "I prefer the term metatext" (p. 121) when commenting on his earlier claim of thesis writers' "greater use of metadiscourse" (Swales, 1990, p. 188). Bunton (1998) did not examine interpersonal items in his metadiscourse analysis; thereby noting that "choice of the term 'metatext' rather than metadiscourse seems particularly apt" (p. 219). Paltridge and Starfield (2007) used metadiscourse and metatext interchangeably, which can be seen from their definition that "meta-discourse (also referred to as metatext) refers to discourse about discourse" (p. 89).

Therefore, this paper addresses a need to examine whether and why research writers make greater use of metadiscourse in their PhD thesis introductions than in their RA introductions based on the definition that metadiscourse consists of both metatextual and interpersonal items.

2.2. Data selection

The corpus analysed in this paper consists of the introductions of PhD theses and RAs by the same authors. They are a part of the data analysed in my previous study (Kawase, 2011) that investigated Japanese researchers' development of academic literacy in English. The theses were selected from those successfully completed at Anglophone (i.e., Australian) universities.

Only RAs published in international journals were chosen. Despite the selection of non-native English-speaking (NNES) writers' texts, this design will allow this paper to explore potential features of the thesis and RA genres shaped by the international academic community where English is used by multicultural populations as the dominant language (cf. Bunton, 1998, 2002).

Following the observation by Connor and Moreno (2005) that selecting comparative texts for analysis is a "vital component in contrastive rhetoric research" (p. 153), the present research selected RAs constructed based on a part or the whole of the theses to ensure comparability of the content as a key determiner of the rhetorical features (e.g., Peacock, 2002; Vassileva, 2000). It turned out that the selected RAs are the authors' initial major publications. This feature of the data will allow the analysis of how writers at a similar stage of their career trajectories and thus at a similar stage of their academic literacy development construct metadiscourse in response to genre expectations. The similarity of the subject matter between theses and RAs by each author and their near-simultaneous production can be seen in Table 1 (see the Appendix for theses and RAs from which data were selected).

Above theses and RAs were constructed in the applied linguistics discipline. The authors of these texts are language specialists considered to be more familiar than researchers in other disciplines with the nature of these research genres (e.g., linguistic and rhetorical features). Hence, this design will provide more opportunities for the present analysis to identify features of metadiscourse that can be ascribed to the genre; if such features exist.

The introductions of the selected texts were identified based on the following criteria. Since the overall structure of the selected RAs basically corresponds to Introduction-Methodology-Results-Discussion (I-M-R-D) structure (e.g., Bazerman, 1988; Swales, 1990), the opening section(s) before the Methods section was identified as the introduction. In selecting PhD thesis introductions, this paper adopted the notion that the entire opening chapter before the Literature Review chapter constitutes the introduction in the case of PhD theses (e.g., Bunton, 1998; Kwan, 2006).

2.3. Analysis

This paper employs Hyland (2005) model as an established framework for the analysis of metatextual and interpersonal items of metadiscourse in academic writing (e.g., Bruce, 2009; Hyland, 2010; Zarei & Mansoori, 2011). One of the innovations of Hyland's model seems to be that it classifies citation as a part of metadiscourse (cf. Adel, 2006; Mauranen, 1993). According to Hyland, this is a necessary move especially for the analysis of metadiscourse in research writings because:

For readers of research papers, claims are inseparable from their originators and a great deal of explicit intertextuality is required from authors to show who first made the claim and how it relates to the current argument. More than this, however, citations are also crucial to gaining approval of new claims by providing persuasive support for arguments and demonstrating the novelty of assertions. (Hyland, 2005, p. 106)

The importance of citation for the texts analysed in this study is thus strongly suggested, given that one of the main rhetorical functions of the introduction of research theses or articles is to justify the research by showing how it is to extend previous research or fill gaps in previous literature (e.g., Paltridge & Starfield, 2007; Samraj, 2008; Swales, 1990; Swales & Feak, 2012).

Table 1

Titles of theses and research articles from which data were selected.

Author1 Text

Nakane(A1) Thesis

Article

Ohashi (A2) Thesis

Article

Iwashita (A3) Thesis

Article

Takeuchi Thesis

Article

Tsukada(A5) Thesis

Article

Hashimoto Thesis

(A6) Article

Machida Thesis

(A7) Article

Hasada (A8) Thesis

Article

Text Title (year of completion/publication)

Silence in Japanese-Australian classroom interaction: Perception and performance (2003) Silence and politeness in intercultural communication in university seminars (2006) Giving, receiving and thanking in Japanese: A cross-cultural pragmatic investigation (2001)

Japanese culture specific face and politeness orientation: A pragmatic investigation of yoroshiku onegaishimasu (2003) The role of task-based conversation in the acquisition of Japanese grammar and vocabulary (1999)

The effect of learner proficiency on interactional moves and modified output in nonnative—nonnative interaction in Japanese as a foreign language (2001)

Raising children bilingually through the 'one parent-one language' approach: A case study of Japanese mothers in the Australian context(2005)

The Japanese language development of children through the 'one parent-one language' approach in Melbourne (2006) An acoustic phonetic analysis of Japanese-accented English (1999)

Cross-language perception of word-final stops by multilingual listeners: Preliminary results on the effect of listeners' first language (L1) backgrounds (2007)

Power and control: Cultural assumptions in Japanese policies on the teaching of Japanese (1997) 'Internationalisation' is 'Japanisation': Japan's foreign language education and national identity (2000) A case study of interlanguage: Japanese language acquisition by English speakers (1996) Japanese text comprehension by Chinese and non-Chinese background learners (2001) An exploratory study of expression of emotions in Japanese: Towards a semantic interpretation (2000) 'Body part' terms and emotion in Japanese (2002)

The above authors are referred to below as A1—A8: She and her are used as generic pronouns to refer to these authors.

In Hyland's model, metatextual items are referred to as "interactive metadiscourse," while interpersonal items are categorised as "interactional metadiscourse." Interactive items (see Table 2) are used to "organise propositional information in ways that a projected target audience is likely to find coherent and convincing" (Hyland, 2005, p. 50).

Interactional items (see Table 3) are used to show "the author's perspective towards both propositional information and readers themselves" (Hyland, 2005, p. 52).

After identifying metadiscourse items based on the above taxonomy, the analysis examines these items from quantitative and qualitative perspectives. The quantitative analyses examine the total number of metadiscourse items employed in each text and the number of occurrences of interactive/interpersonal items as well as individual metadiscourse items. The qualitative analyses examine how metadiscourse items are used for different purposes and effects to consider possible reasons for the variations.

3. Results and discussion

3.1. Quantitative approach

The analysis shows that six out of the eight writers make greater use of metadiscourse in their RA introductions than in their thesis introductions; see Fig. 1 summarising the total number of metadiscourse items used in these texts:

The analysis of interactive resources (Fig. 2) shows that six writers make greater use of metatextual items in their RAs. These results are inconsistent with the assumption that thesis writers make greater use of metadiscourse in general and metatextual items in particular (e.g., Bunton, 1998; Paltridge & Starfield, 2007; Swales, 1990, 2004). The analysis of interactional items shows that four writers use more interpersonal items in their RAs; see Fig. 3:

It is thus revealed that among the six authors who made greater total use of metadiscourse in their RA introductions, the majority (i.e., A1, A2, A3, and A7) used both interactive and interpersonal items more in the introductions. This finding is novel, considering that previous studies tend to focus on metatextual items alone when discussing differences in the overall quantity of metadiscourse in PhD theses and RAs (cf. Swales, 2004).

The analysis of individual items shows that among the changes made by the majority of the writers are greater use of evidentials, code glosses, and hedges in RAs. Items such as transition markers, endophoric markers, attitude markers, and self-mention, in contrast, tend to be used less in RAs. In other words, the authors who made greater (or lesser) total use of metatextual or interpersonal resources does not make greater (or lesser) use of all subtypes; see Fig. 4 and Fig. 5:

It can be thus said that differences in overall quantity of metatextual or interpersonal items cannot be explained in a reductive fashion; such as that the overall greater use of metatextual items suggests the writers' greater efforts to explicitly guide readers through text (cf. Bunton, 1998). Seen in this light, greater (or lesser) total use of metadiscourse cannot be accounted for in such a fashion, either, provided especially that most of the above mentioned changes in the use of individual items are also observed in the introductions by A6 and A8 making less overall use of metadiscourse in their RA introductions. The following sections take a closer look at how metadiscourse items are used in specific contexts to further discuss the implications of the changes.

3.2. Qualitative approach

3.2.1. Interactive resources: endophoric markers, evidentials, code glosses, and transition markers

The analysis shows that although all authors except A6 use endophoric markers to present chapter previews in their thesis introductions, none of them do so in their RA introductions. Such markers can be seen in the final paragraph of A2's thesis introduction, to give an example; see Ex. 1, where endophoric markers are underlined:

Ex. 1 Chapter 1 has provided a general overview of the study. In Chapter 2, theoretical issues which are significant to the study are discussed. Several controversial issues that remain unsolved are highlighted. The theoretical discussion includes pragmatics (2.1), speech act theory (2.2), conversation analysis (2.3), the cooperative principle (2.4), indirect speech acts (25) and politeness theory (26). In Chapter 3, merits and demerits of data elicitation methods such as DCTs and role plays are discussed. This is followed by Study 1 and Study 2. Japanese native speakers' common linguistic expressions, sensitivity to social debt, and sequential patterns (such as 'thanking — denial') are identified for further investigation. In Chapter 4, Study 3 investigates multiple speech act (requesting, thanking and declining) realisations in a series of latter writing tasks. Letters of different first language backgrounds (Chinese and English) and achievement

Table 2

Interactive metadiscourse defined by Hyland (2005).

Metadiscourse Functions Examples

Transitions Frame markers Endophoric markers Evidentials Code glosses Express semantic relation between main clauses or sentences Refer to discourse acts, sequences or text stages Refer to information in other part of the text Refer to sources of information from other texts Help readers grasp meanings of ideational material In addition/but/and Finally/my purpose is to Noted above/in Section2 According to X/(Y, 1990) Namely/e.g./such as/In other words

Table 3

Interactional metadiscourse defined by Hyland (2005).

Metadiscourse

Functions

Examples

Hedges Boosters

Attitude markers Engagement markers Self-mention

Withhold commitment to a proposition and open dialogue Emphasise certainty or close dialogue Express writer's attitude to proposition

Explicitly build relationship with the reader by addressing to the reader Explicit reference to author(s)

Might/perhaps/possible/about In fact/definitely/it is clear that Unfortunately/surprisingly Consider/note that/you can see that I/We/my/our

levels (high and low) are involved, as well as Japanese native speakers. Native speaker judgement on learners' letter writing tasks is used to examine this data. This reveals the native speaker's cultural values and what is perceived as polite behaviour. In Chapter 5, Study 4 investigates Japanese native speakers' speech act realisations of thanking in naturally occurring telephone conversations. Chapter 6 discusses the outcomes of these studies in the light of the theoretical framework presented in Chapter 2, and also discusses the implications and limitations of the research outcome. (A2's thesis, sentences 121—133)

Considering that presenting chapter summaries has been often indicated as a mandatory element of PhD thesis introductions (e.g., Bunton, 1998; Paltridge & Starfield, 2007), greater use of endophoric markers for this reason can be attributed to the genre.

The present analysis identifies that the majority of the authors (i.e., A1, A2, A3, A4, A5, and A6), in contrast, use evidentials less in their thesis introductions. Close analysis reveals that this feature arises because these authors do not always give comprehensive reports of the literature when they refer to previous research in their thesis introductions. Such citation practice, for instance, can be seen in the following excerpt from A1's thesis introduction (Ex. 2, an underlined item is evidential):

Ex. 2 As discussed in Chapter 2, the existing literature tends to rely on personal observations and anecdotal accounts from colleagues or students to account for Asian students' silence (Liu 2000). (A1's thesis, sentence 31)

As is noticeable, the above evidential does not represent a part of "the existing literature" but is used merely to support her comment on the methodological trend. It can be also inferred from an endophoric expression ("As discussed in Chapter 2") that A1 was allowed not to give comprehensive reports of the literature until the Literature Review chapter (e.g., Kwan, 2006). Such deferral can be considered an acceptable feature in the PhD thesis genre; otherwise, texts like the opening paragraph of A3's thesis introduction (see Ex. 3 below), for example, would not have been possible:

Ex. 3 In communicative language teaching, a widely held belief is that the more opportunities learners have for communication with native speakers or with non-native speakers of high proficiency, the more competent learners will become in the foreign language. Thus, many foreign language educators make efforts to create opportunities for learners to use the target language, for example, by inviting native speakers into the classroom: pairing up foreign language learners with volunteer or paid partners who are native speakers of that language; organising 'tandem' exchanges and including a study abroad program as a part of the curriculum. A question which arises is whether increasing these opportunities for learners, especially beginners to converse with native speakers is really helpful for language learning. Although there are theoretical grounds for believing that interactions with native speakers create conditions that are facilitative for second language acquisition, there is a need to investigate this empirically, both cross-sectionally and longitudinally with real learners conversing with native speaking partners, using a variety of topics and tasks. The present study investigates the role of conversation in the acquisition of Japanese grammar and vocabulary. It examines the effect of task-based conversation in both the short- and long-term. (A3's thesis, sentences

A3 does not employ a single citation to support her statements in this paragraph, although some of them (e.g., a comment on the "theoretical grounds") would clearly require evidential support. The assumption that this feature is genre-specific can be further supported by the finding that A3 uses evidentials in the opening paragraph of her RA introduction (see Ex. 4 below, evidentials are underlined), which is strikingly similar in content.

Ex. 4 In communicative language classrooms, group work using communication tasks has been widely implemented, and learners spend a great deal of time practising the target language with other learners. Long and Porter (1985) claim that group work enhances opportunities for students to use the target language, and to interact with one another. One aspect of conversational interaction, known as negotiation of meaning, has been claimed to be facilitative in language learning (e.g. Long, 1980,1983; Pica, 1987,1994a, 1994b). Empirical studies have demonstrated that learners have more opportunity to talk and negotiate meaning in group work than in the teacher-fronted classroom (e.g. Long, 1981; Pica and Doughty, 1985; Doughty and Pica, 1986). (A3's RA, sentences 1—4)

As can be seen, A3 establishes within the paragraph that the topic of language learners' benefits from conversational interaction is an acknowledged one in the field.

1—6)

A1 A2 A3 A4 A5 A6 A7 A8

Fig. 1. Number of metadiscourse items per 100 sentences (normalised).

140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

I Thesis RA

A1 A2 A3 A4 A5 A6 A7 A8

Fig. 2. Number of interactive items per 100 sentences (normalised).

140 -13 .7

130 120 110 100 90 80 70 60 50 40 30 20 10 0

A1 A2 A3 A4 A5 A6 A7 A8

Fig. 3. Number of interactional items per 100 sentences (normalised).

I Thesis RA

The analysis suggests that greater use of code glosses in RA introductions, too, may derive from genre-related factors. Close examination reveals that the authors using this item more in RAs (i.e., A1, A2, A3, A5, and A6) tend to use it for describing methodological information (see Ex. 5, code glosses are underlined).

Ex. 5 In this paper, I examine through one policy text, entitled Japanese Government Policies in Education, Science and Culture 1994; New Directions in School Education: Fostering Strength for Life (JGPESC), with some supplementary reference to other years' policies. JGPESC is a report, which has been published by the Ministry of Education, Science and Culture (Monbushou, MESC) annually since 1959, and has a specific subtitle for each year. The JGPESC is located as the final version of the policies produced in that year. The JGPESC is, therefore, the endorsed public position of the MESC. The structure of the JGPESC does not vary a great deal except in the inclusion of the special feature for each year. The 1994 document has been chosen because it specifically features school education. The 1994 document also has the first complete translation in English. This document is referred to as a White Paper which informs 'the public about all aspects of educational administration over the past year' (MESC, 1995, p. xiii). The Japanese original title is Wagakuni no

Fig. 4. Occurrences of interactive items per 100 sentences (normalised). This figure illustrates increase or decrease in the use of Transition markers (TR), Frame markers (FM), Endophoric markers (END), Evidentials (EV), and Code glosses (CG) between Theses (TH) and RAs.

^ V* r r r r ^ rrrrrrrrV^

HG EM SM B

Fig. 5. Occurrences of interactional items per 100 sentences (normalised). This figure illustrates increase or decrease in the use of Hedges (HG), Attitude markers (AM), Self-mention (SM), and Engagement markers (EM) between Theses (TH) and RAs. Please note that the value for SM in A6's thesis is 111.6.

bunkyou shisaku, literally 'cultural and educational policies for our (own) country', and is couched in a very formal style of Japanese. (A6's RA, sentences 58—66)

Type of code glosses seen above (i.e., those used for data description) is also found in the introductions by other four writers. However, such rhetorical practice is an essential element of the Methods section (e.g., Swales, 2004; Swales & Feak, 2012). We should note that, of the I-M-R-D structure of RAs, the Methods section is "the narrowest part" (Swales & Feak, 2012, p. 285) and thus that the writers are least expected to "over-explain" (p. 295). Despite this, RA authors in the field of language studies and teaching usually construct the Methods section that is more elaborate than, for example, those in hard sciences (Swales, 2004). Hence, the language researchers whose works were analysed in this paper might choose to include methodological information in the introductions so that they can downsize the Methods section. While this seems to be an underlying reason for the greater use of code glosses in RA introductions, such an effort is not necessary for thesis authors. This is because, according to Paltridge and Starfield (2007, p. 114), the "lengthier Methodology components of research theses as

opposed to the more terse research article" are meant for doctoral students to demonstrate their ability to make "argument and justification for the chosen approach."

Unlike in the case of the interactive items discussed above, the analysis did not identify textual evidence to account for why the majority of the writers (i.e., A2, A3, A4, A5, and A6) use transition markers (transitions) less in their RA introductions. Nevertheless, it is suggested that this result, at least, cannot be explained in skill-deficit terms, although greater use of transitions itself has been often regarded as an indicator of writers' better skills in English academic writing (e.g., Hyland, 2004). The present analysis has examined metadiscourse in PhD theses and subsequently published RAs produced by the same authors within a short interval. Therefore, it is unlikely that variations in the use of metadiscourse between these texts are due to the difference in the level of the authors' skills. Seen in this light, the following genre-specific features seem to be more relevant. RAs assume "tacit understanding between 'expert' writers and readers and thus tend to be less explicit in their exposition than we might expect in dissertations" (Swales, 2004, p. 119). In contrast, authors of PhD theses, who are only "seeking to enter a community of scholars" (Paltridge & Starfield, 2007, p. 82), would "prefer to exercise a more transparent communicative style on the grounds that it is safer" (Adel, 2006, p. 153). Moreover, in "extended texts such as theses, metadiscourse plays an important role in that it helps provide an overarching organizational scaffold for the thesis" (Paltridge & Starfield, 2007, p. 89).

3.2.2. Interactional resources: attitude markers, self-mention, and hedges

The analysis of interactional metadiscourse (cf. Fig. 5) shows that four writers (i.e., A3, A5, A6, and A8) use attitude markers less in their thesis introductions and that six writers (i.e., A3, A4, A5, A6, A7, and A8) use self-mention less in their thesis introductions. Close examination reveals that decrease in the use of these items is because while these writers use them to comment on previous research in their theses, they use evidentials to do so in their RAs. Some examples will illustrate how this shift may contribute to the tonal difference, followed by discussion of the implications of these changes.

The writers who used attitude markers more frequently in their thesis introductions tend to employ evaluative language characteristic of this metadiscourse item to assert insufficient aspects or gaps in previous research. A resultant personal tone can be found in the following examples (Ex. 6 and Ex. 7) (markers underlined):

Ex. 6 However, we are still far from being able to draw any firm conclusions about the role of negative evidence in L1 language acquisition. (A3's thesis, sentence 57)

Ex. 7 It is somewhat unrealistic and most regrettable that "foreign accent" was regarded by some researchers as a stigma. (A5's thesis, sentence 9)

These writers, by contrast, use evidentials in their RA introductions to indicate insufficiencies in previous knowledge; see Ex. 8 and Ex. 9 (evidentials underlined):

Ex. 8 To date, however, studies on modified output have only concerned the gender variables (Pica et al., 1989,1991; Pica, 1991) in NS-NNS interactions. (A3's RA, sentence 44)

Ex. 9 However, not all foreign speech sound contrasts are equally difficult to perceive and/or produce (e.g., voicing vs. palate contrast in Hindi, Tees & Werker, 1994). (A5's RA, sentence 3)

As can be seen, these writers use evidentials to describe the status of research knowledge in a more detached manner through other studies.

The difference in tone due to the shift from self-mention to evidentials is most clearly seen in A4's introductions, to give an example, where she presents seemingly identical research (see Ex. 10 and Ex. 11; self-mention and evidentials are underlined).

Ex. 10 At the moment, our knowledge of bilingual language development through the 'one parent-one language' approach is limited to young children. The period around when a child starts schooling in the majority language seems to be a crucial point which can lead either to language maintenance or language shift. We need to know more about what happens to the child at this time. (A4's thesis, sentences 9—11)

Ex. 11 Studies on the 'one parent-one language' approach reported varying outcomes (Amberg, 1987; Billings, 1990; Do; pke, 1992; Harding & Riley, 1986; Lanza, 1997; Leopold, 1939—1949; Noguchi, 1996; Ronjat, 1913; Sanders, 1982,1988; Taeschner, 1983; Yamamoto, 1995). The 'one parent-one language' approach is understood to be relatively effective in promoting active bilingualism among young children. When children approach school age, input in the majority language increases. This period seems to be a crucial point where some children continue to speak in the minority language, whilst other children shift towards the majority language. However, not enough is known about the minority language development among older children. (A4's RA, sentence 4—8)

In the thesis, A4 highlights her authorial presence by using we and our to assert the status of knowledge of the "one parent-one language approach" as a part of those who represent the field. In contrast, in the RA, A4 achieves an impersonal tone by using evidentials to describe the information about the approach as previous findings.

The present study interprets variations in the use of the above interactional resources by arguing that more personal constructions in PhD thesis introductions are closely related to the following genre-specific features. According to Kamler and

Thomson (2006), doctoral students would initially find it difficult to present an authoritative self in writing when assessing previous studies; because of their increasing awareness that these studies are produced by more authoritative members in the research field (see also Koutsantoni, 2006,2007). As a result, some of these apprentices could turn to focus on taking notes (e.g., literature summaries) and miss the gist of their own project in the process, which could lead to a delay in the progress of thesis writing. Partly because of this, doctoral students are often advised by their supervisors that it is "the personal stamp, the angle that students take on their research problem, that really makes their contribution to knowledge original" (Kamler & Thomson, 2006, p. 73; see also Paltridge & Starfield, 2007). This seems to be why it is common for PhD thesis authors to refer to their personal contexts (e.g., those related to their professional experiences) when they justify their research in their introductions (e.g., McNabb, 2001). Accordingly, thesis authors' metadiscoursal manifestation of their authorial presence when commenting on previous research, too, can be ascribed to this feature of PhD theses as a more educational genre functioning as "a final examination in a long student career" (Johns & Swales, 2002, p. 16) conducted in an apprenticeship environment (e.g., Paltridge & Starfield, 2007).

We should also note that according to previous research, using attitude markers and self-mention (we/our) to describe insufficient aspects of previous knowledge could be an acceptable method in research genres. Although using attitude markers to evaluate previous studies would create unintended antagonisms and thus is undesirable (e.g., Lewin, Fine, & Young, 2001) especially in a "high-visibility" section like the introduction (Canagarajah, 2002, p. 179), this could be deemed acceptable when used with hedges to redress such threatening effects (e.g., Myers, 1989). In fact, the writers whose works were analysed in this paper used attitude markers along with hedges in their thesis introductions to describe gaps in previous research ("still" and "any" in Ex. 6 and "somewhat" in Ex. 7). Moreover, as Hyland (2005) pointed out, it is a fairy common practice, especially for researchers in the humanities disciplines, to indicate "recognized gaps in existing knowledge" (p. 150) by using attitude markers. Likewise, using self-mention (we/our) to comment on previous research in the introductions has been indicated as a conventional method in academic genres (e.g., Harwood, 2005; Kuo, 1999). According to Swales (1990), the authors of the introductions of research papers could use we to present themselves "as co-members of the unsuccessful group" (p. 156) when describing gaps in previous research. Myers (1989) also noted that writers can use we to "include themselves in the criticism" (p. 7) to redress its potentially threatening effects.

The following question arises: Why did the writers stop using these items to comment on previous knowledge, provided that this can be acceptable in RA introductions? We should note here that the RAs analysed in this paper were produced by the authors as prospective newcomers in the academic publishing community (cf. Koutsantoni, 2007). At this stage of their career, they might have become heartily aware that their "relationship to a more authoritative community of scholars" was still "uncertain or in formation" (Kamler & Thomson, 2006, p. 60; see also Johns & Swales, 2002; Paltridge & Starfield, 2007; Swales, 2004). Moreover, they must have faced the reality that they need to survive "the fierce academic competition to get their papers published" (Paltridge & Starfield, 2007, P. 82; see also Casanave & Vandrick, 2003). In fact, it is not rare for such novices to have their article manuscripts rejected, even though these manuscripts usually go through extensive revising and editing before and after being submitted to the targeted journals (e.g., Burrough-Boenisch, 2003; Curry & Lillis, 2004; Flowerdew, 2000; Li & Flowerdew, 2007). Kubota (2003), for instance, described a possible consequence of this in the following terms:

In the process of revising the paper, however, I increasingly felt as though I was writing my paper on behalf of the reviewer.... My occupation to have the paper accepted compelled me to comply with every single suggestion given by the reviewer.... I eventually felt that much of the content and language did not really belong to me; I felt I had lost the ownership of my ideas and words. (p. 63)

The shift from attitude markers and self-mention (we/our) to evidentials, hence, appears to be a result of the writers' efforts to construct metadiscourse based on the deeper awareness of their dependent position in the academic publishing enterprise. They might seek to avoid any potentially risky elements for a better chance to get their papers published by conforming to more standard methods of the RA genre constituting an "elaborate system of knowledge production" (Myers, 1989, p. 25) where authors are generally expected to be experts with an impersonal stance (e.g., Bazerman, 1988). Furthermore, this decision seems to derive from their understanding of a genre convention that the "[introduction is extremely important in positioning the writer as having something to say that is worth publishing" (Paltridge & Starfield, 2007, P. 82).

The analysis also shows that the above mentioned preference of evidentials in RA introductions may lead to greater use of hedges. As has been suggested (e.g., Lewin et al., 2001; Swales, 1990), writers of RA introductions would use hedges most when commenting about or from previous research referred to through evidentials. The present analysis, in fact, identifies that the combined use of these items is recurrently employed in the RA introductions by four writers (i.e., A1, A2, A3, and A4) who make greater use of both items in their RAs than in their theses (cf. Figs. 4 and 5). The combination is shown in Ex. 8 presented earlier ("To date" is a hedge) and Ex. 12 in the following with hedges and evidentials underlined:

Ex. 12 Considering that the Japanese are often described as attaching strong values to and making abundant use of silence (e.g. Branlund, 1975; Loveday, 1982; La Forge, 1983; Lebra, 1987) and assuming that Australian culture tilts towards American or British culture, Australian-Japanese intercultural communication may be expected to exhibit positive and negative politeness as well as a talk-silence contrast. (A1's RA introduction, sentence 18)

As can be seen from the examples, writers of the introductions would use hedges along with evidentials to describe previous knowledge in a less assertive manner so that they can "convey deference and respect for readers' views" (Hyland,

2005, p. 68) to involve them in the process of contextualising the study (e.g., Lewin et al., 2001; McNabb, 2001). Therefore, provided that greater use of evidentials in RAs is a genre-derived feature, a resultant greater hedging can also be ascribed to the genre.

4. Concluding remarks

It is inherently difficult to infer the production processes from the textual product alone. Absence of interviews with the authors of the selected texts can thus be indicated as one of the limitations of the present research. Considering also the size and scope of the research, any conclusions drawn from the findings will require further exploration. However, the present research suggests that the claim that research writers make greater use of metadiscourse as well as metatext in PhD theses than in RAs may not be applicable with regard to the introductions. Moreover, variations in the use of metadiscourse in PhD thesis introductions and RA introductions may derive from features specific to the PhD thesis genre or the RA genre. Noteworthy findings were also that changes in the use of some interactive and interactional items can be due to similar reasons (e.g., related to how research writers construct their authorial presence in their theses or RAs) and that increased use of an interactive item may be an underlying factor of greater use of a relevant interactional item. These findings suggest that the construction of interactive items and interactional items can be closely related. Furthermore, some of the writers did not use specific items in RAs in the way they did in their theses, even if their use was not discordant with the conventions of research writing. This might shed further light on complex rhetorical manoeuvres expected in the research genres.

Currently, novice and/or NNES researchers are often advised of greater use of metadiscourse in their academic writing (e.g., Bunton, 1998; Paltridge & Starfield, 2007). Nevertheless, as shown in this paper, greater or less use of metadiscourse items with particular functions can be a result of writers' rhetorical strategies adopted in response to the nature of the section of the targeted genre. Therefore, it would be useful for novices, especially NNES writers from rhetorical traditions in which metadiscoursal guidance of the reader is not a conventional feature, to learn not only the rhetorical functions of meta-discourse items but also how they should be orchestrated to construct their desired rhetoric.

Appendix

Theses and research articles from which data were selected:

Hasada, R. (2000). An exploratory study of expression of emotions in Japanese: Towards a semantic interpretation (Unpublished doctoral thesis). Australian National University, Australia.

Hasada, R. (2002). 'Body part' terms and emotion in Japanese. Pragmatics and Cognition, 10(1), 107—128.

Hashimoto, K. (1997). Power and control: cultural assumptions in Japanese policies on the teaching of English (Unpublished doctoral thesis). La Trobe University, Australia.

Hashimoto, K. (2000). 'Internationalisation' is 'Japanisation': Japan's foreign language education and national identity. Journal of Intercultural Studies, 21(1), 39—51.

Iwashita, N. (1999). The role of task-based conversation in the acquisition of Japanese grammar and vocabulary (Unpublished doctoral thesis). University of Melbourne, Australia.

Iwashita, N. (2001). The effect of learner proficiency on interactional moves and modified output in nonnative—nonnative interaction in Japanese as a foreign language. System, 29, 267—287.

Machida, S. (1996). A case study of interlanguage: Japanese language acquisition by English speakers (Unpublished doctoral thesis). University of Queensland, Australia.

Machida, S. (2001). Japanese text comprehension by Chinese and non-Chinese background learners. System, 29,103—118.

Nakane, I. (2003). Silence in Japanese-Australian classroom interaction: Perception and performance (Doctoral thesis, University of Sydney, Australia). Retrieved from http://ses.library.usyd.edu.au/handle/2123/568.

Nakane, I. (2006). Silence and politeness in intercultural communication in university seminars. Journal of Pragmatics, 38(11), 1811—1835.

Ohashi, J. (2001). Giving, receiving and thanking in Japanese: A cross-cultural pragmatic investigation (Unpublished doctoral thesis). University of Melbourne, Australia.

Ohashi, J. (2003). Japanese culture specific face and politeness orientation: A pragmatic investigation of yoroshiku one-gaishimasu. Multilingual Journal of Cross Cultural and Interlanguage Communication, 22, 257—274.

Takeuchi, M. (2005). Raising children bilingually through the 'one parent-one language' approach: A case study of Japanese mothers in the Australian contexts (Unpublished doctoral thesis). Monash University, Australia.

Takeuchi, M. (2006). The Japanese language development of children through the 'one parent-one language' approach in Melbourne. Journal of Multilingual and Multicultural Development, 27(4), 319—331.

Tsukada, K. (1999). An acoustic phonetic analysis of Japanese-accented English (Unpublished doctoral thesis). Macquarie University, Australia.

Tsukada, K. (2007). Cross-language perception of word-final stops by multilingual listeners: preliminary results on the effect of listeners' first language (L1) background. Asia Pacific Journal of Speech, Language, and Hearing, 10(1), 1—13.

References

Adel, A. (2006). Metadiscourse in L1 and L2 English. Amsterdam/Philadelphia: John Benjamins.

Bazerman, C. (1988). Shaping written knowledge: The genre and activity of the experimental article in science. Madison, WI: University of Wisconsin Press.

Bruce, I. (2009). Results sections in sociology and organic chemistry articles: a genre analysis. English for Specific Purposes, 28(2), 105—124.

Bunton, D. (1998). Linguistic and textual problems. Ph.D. and M.Phil. theses: An analysis of genre moves and metatext (Doctoral thesis. Hong Kong: University of Hong Kong. Retrieved from http://hub.hku.hk/handle/10722/39580.

Bunton, D. (2002). Genre moves in Ph.D. thesis introductions. In J. Flowerdew (Ed.), Academic discourse (pp. 57—75). Harlow: Longman.

Burrough-Boenisch, J. (2003). Shapers of published NNS research articles. Journal of Second Language Writing, 12(3), 223—243.

Canagarajah, A. S. (2002). Critical academic writing and multilingual students. Ann Arbour, MI: University of Michigan Press.

Casanave, C. P., & Vandrick, S. (2003). Writing for scholarly publication: Behind the scenes in language education. Mahwah, NJ: Lawrence Erlbaum.

Connor, U. M., & Moreno, A. I. (2005). Tertium comparationis: a vital component in contrastive rhetoric research. In P. Bruthiaux, D. Atkinson, W. G. Eggington, W. Grabe, & V. Ramanathan (Eds.), Directions in applied linguistics: Essays in honor of Robert B. Kaplan (pp. 153—164). Clevedon, England: Multilingual Matters.

Crismore, A. (1989). Talking with readers: Metadiscourse as rhetorical act. New York: Peter Lang<x>.

Crismore, A., Markkanen, R., & Steffensen, M. S. (1993). Metadiscourse in persuasive writing. Written Communication, 10(1), 39—71.

Curry, M. J., & Lillis, T. (2004). Multilingual scholars and the imperative to publish in English: negotiating interests, demands, and rewards. TESOL Quarterly, 38(4), 663—688.

Feak, C. B., Swales, J., & Irwin, V. (2011). Online commentary to creating contexts: Writing introductions across genres. Retrieved from http://www.press.umich. edu/script/press/elt/compsite/ETRW/9780472034567-commentary.pdf.

Flowerdew, J. (2000). Discourse community, legitimate peripheral participation, and the nonnative-English-speaking scholar. TESOL Quarterly, 34(1), 127—150.

Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Edward Arnold.

Harwood, N. (2005). 'We do not seem to have a theory ... The theory I present here attempts to fill this gap': inclusive and exclusive pronouns in academic writing. Applied Linguistics, 26(3), 343—375.

Hyland, K. (2000). Disciplinary discourses: Social interactions in academic writing. London: Longman.

Hyland, K. (2004). Disciplinary interactions: metadiscourse in L2 postgraduate writing. Journal of Second Language Writing, 13,133—151.

Hyland, K. (2005). Metadiscourse: Exploring interaction in writing. London: Continuum.

Hyland, K. (2010). Metadiscourse: mapping interactions in academic writing. Nordic Journal of English Studies., 9(2), 125—143.

Hyland, K., & Tse, P. (2004). Metadiscourse in academic writing: a reappraisal. Applied Linguistics, 25(2), 156—177.

Johns, A. M., & Swales, J. M. (2002). Literacy and disciplinary practices: opening and closing perspectives. Journal of English for Academic Purposes, 1(1), 13—28.

Kamler, B., & Thomson, P. (2006). Helping doctoral students write: Pedagogies for supervision. New York: Routledge.

Kawase, T. (2011). Japanese novice researchers' development of rhetorical control in academic English. Unpublished doctoral thesis. Australia: Deakin University.

Koutsantoni, D. (2006). Rhetorical strategies in engineering research articles and research theses: advanced academic literacy and relations of power. Journal of English for Academic Purposes, 5(1), 19—36.

Koutsantoni, D. (2007). Developing academic literacies: Understanding disciplinary communities' culture and rhetoric. Oxford: Peter Lang<x>.

Kubota, R. (2003). Striving for original voice in publication?: a critical response. In C. P. Casanave, & S. Vandrick (Eds.), Writing for scholarly publication: Behind the scenes in language education (pp. 61—72). Mahwah, N.J: Lawrence Erlbaum.

Kuo, C. H. (1999). The use of personal pronouns: role relationships in scientific journal articles. English for Specific Purposes, 18(2), 121—138.

Kwan, B. S. (2006). The schematic structure of literature reviews in doctoral theses of applied linguistics. English for Specific Purposes, 25(1), 30—55.

Lewin, B. A., Fine, J., & Young, L. (2001). Expository discourse: A genre-based approach to social science research text. New York: Continuum.

Li, Y., & Flowerdew, J. (2007). Shaping Chinese novice scientists' manuscripts for publication. Journal of Second Language Writing, 16(2), 100—117.

Mauranen, A. (1993). Contrastive ESP rhetoric: metatext in Finnish-English economics texts. English for Specific Purposes, 12(1), 3—22.

McNabb, R. (2001). Making the gesture: graduate student submissions and the expectation of journal referees. Composition Studies, 29(1), 9—26.

Myers, G. (1989). The pragmatics of politeness in scientific articles. Applied Linguistics, 10(1), 1—35.

Paltridge, B., & Starfield, S. (2007). Thesis and dissertation writing in a second language: A handbook for supervisors. London: Routledge.

Peacock, M. (2002). Communicative moves in the discussion section of research articles. System, 30(4), 479—49 .

Samraj, B. (2008). A discourse analysis of master's theses across disciplines with a focus on introductions. Journal of English for Academic Purposes, 7(1), 55—67.

Shaw, P. (2000). Towards classifying the arguments in research genres. In A. Trosborg (Ed.), Analysing professional genres (pp. 41—56). Philadelphia, PA: John Benjamins.

Swales, J. (1990). Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press.

Swales, J. (2004). Research genres: Explorations and applications. New York: Cambridge University Press.

Swales, J., & Feak, C. B. (2012). Academic writing for graduate students: Essential tasks and skills (3rd ed.). Ann Arbor, MI: University of Michigan Press.

Vande Kopple, W. (1985). Some exploratory discourse on metadiscourse. College Composition and Communication, 36, 82—93.

Vande Kopple, W. (2002). Metadiscourse, discourse and issues in composition and rhetoric. In E. Barton, & G. Stygall (Eds.), Discourse studies in composition (pp. 91—113). Cresskill, NJ: Hampton Press.

Vassileva, I. (2000). Who is the author?: A contrastive analysis of authorial presence in English, German, French and Bulgarian academic discourse. Sankt Augustin, Germany: Asgard Verlag.

Zarei, G. R., & Mansoori, S. (2011). A contrastive study on metadiscourse elements used in humanities vs. non humanities across Persian and English. English Language Teaching, 4(1), 42—50.