The use of conjunctive and lexical relations for elabaration in expository writing by intermediate level efl learners ma

Motivated by Halliday and Matthiessen‘s elaboration 2014, this study focused on conjunctions and lexical relations for elaborating relations in EFL expository writing.. The results contr

VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY

UNIVERSITY OF SOCIAL SCIENCES AND HUMANITIES

FACULTY OF ENGLISH LINGUISTICS AND LITERATURE

THE USE OF CONJUNCTIVE AND LEXICAL RELATIONS FOR ELABORATION IN EXPOSITORY

WRITING BY INTERMEDIATE-LEVEL EFL

DANG TAN TIN, Ph.D

HO CHI MINH CITY, OCTOBER 2021

VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY

UNIVERSITY OF SOCIAL SCIENCES AND HUMANITIES

FACULTY OF ENGLISH LINGUISTICS AND LITERATURE

THE USE OF CONJUNCTIVE AND LEXICAL RELATIONS FOR ELABORATION IN EXPOSITORY

WRITING BY INTERMEDIATE-LEVEL EFL

DANG TAN TIN, Ph.D

HO CHI MINH CITY, OCTOBER 2021

ACKNOWLEDGEMENTS

I would like to express my gratitude to those who have helped and supported me during the hard time doing my research and completing this thesis First, my gratefulness goes to my respectful supervisor, Dr Dang Tan Tin, for encouraging me

to continuously review literature, write up and revise my thesis and providing me with valuable feedback and precious guidance throughout the tiring stages of doing research and thesis writing

I am thankful for the continuous support that the management of the Faculty of English Linguistics and Literature provided when I was working on my thesis My genuine appreciation also goes to my colleagues, for all the generous and enthusiastic support as well as the invaluable and professional advice they have given to help me better understand the issue under investigation

Last but not least, I am very grateful to my incomparably caring and loving family for always being by my side to cheer me up when I suffered stress from my work and study Without the encouragement, support and guidance provided by those named or unlisted, I would not have completed this Master‘s Thesis

I appreciate all

STATEMENT OF ORIGINALITY

I hereby certify my authorship of the thesis submitted today entitled:

THE USE OF CONJUNCTIVE AND LEXICAL RELATIONS FOR

ELABORATION IN EXPOSITORY WRITING BY

INTERMEDIATE-LEVEL EFL LEARNERS

In terms of the statement of Requirements for the Thesis in Master‘s Program issued

by the Higher Degree Committee The thesis has not been submitted for the award of any degree or diploma in any other situations

Ho Chi Minh City, Octorber 2021

Nguyen Hoang Kham

RETENTION AND USE OF THE THESIS

I hereby state that I, Nguyen Hoang Kham, being the candidate for the degree of Master in TESOL, accept the requirements of University of Social Sciences and Humanities relating to the retention and use of Master‘s Theses deposited in the library

In terms of these conditions, I agree that the original of my thesis deposited in the library should be accessible for the purpose of study and research, in accordance with the normal conditions established by the library for the care, loan or reproduction of the thesis

Ho Chi Minh City, October 2021

Nguyen Hoang Kham

TABLE OF CONTENTS

ACKNOWLEDGEMENTS i

STATEMENT OF ORIGINALITY ii

RETENTION AND USE OF THE THESIS iii

TABLE OF CONTENTS iv

LIST OF ABBREVIATIONS vii

LIST OF TABLES viii

LIST OF FIGURES ix

ABSTRACT 1

CHAPTER 1 INTRODUCTION 2

1.1 BACKGROUND TO THE STUDY 2

1.2 AIM OF THE STUDY 4

1.3 RESEARCH QUESTIONS 5

1.4 SCOPE OF THE STUDY 5

1.5 SIGNIFICANCE OF THE STUDY 6

1.6 ORGANIZATION OF THE THESIS 6

CHAPTER 2 LITERATURE REVIEW 7

2.1 EXPOSITORY WRITING 7

2.2 HALLIDAY AND MATTHIESSEN‘S MODEL OF ELABORATION 10

2.2.1 Terminology 10

2.2.2 Categories of elaboration 14

2.3 CONJUNCTIVE RELATIONS FOR ELABORATION 15

2.3.1 Terminology 16

2.3.2 Categories of conjunctive relations 18

2.4 LEXICAL RELATIONS FOR ELABORATION 21

2.4.1 Terminology 21

2.4.2 Categories of lexical relations 22

2.4.3 Lexical relations and context 27

2.5 THE FRAMEWORK OF ANALYSIS IN THE PRESENT STUDY 30

2.6 SUMMARY 31

CHAPTER 3 METHODOLOGY 32

3.1 RESEARCH DESIGN 32

3.2 RESEARCH SITE 33

3.3 PARTICIPANTS 34

3.4 ETHICS 35

3.5 WRITING TASK 35

3.6 PROCEDURE OF DATA COLLECTION 37

3.7 DATA ANALYSIS PROCEDURE 38

3.7.1 Segmentation 38

3.7.2 Identification of topic sentence 39

3.7.3 Analysis of elaboration 41

3.7.4 Analysis of conjunctive relations 43

3.7.5 Analysis of lexical relations 44

3.7.6 Statistical analysis 47

3.8 INTER-RATER RELIABILITY 48

3.9 SUMMARY 49

CHAPTER 4 RESULTS AND DISCUSSIONS 49

4.1 RESULTS 50

4.1.1 Elaboration 50

4.1.2 Conjunctive relations for elaboration 52

4.1.3 Lexical relations for elaboration 54

4.2 DISCUSSION 55

4.2.1 Elaboration 55

4.2.2 Conjunctive relations for elaboration 57

4.2.3 Lexical relations for elaboration 58

CHAPTER 5 CONCLUSIONS 61

5.1 SUMMARY OF RESULTS 61

5.2 LIMITATIONS OF THE STUDY 62

5.3 IMPLICATIONS 63

5.3.1 Writing instruction 63

5.3.2 Future research 65

REFERENCES 66

APPENDIX A 71

APPENDIX B 74

APPENDIX C 75

APPENDIX D 76

LIST OF ABBREVIATIONS

EFL English as a Foreign Language

ELT English language teaching

IB IELTS basic (the name of the IELTS course under study)

IELTS International language testing system

TOEFL Test of English as a foreign language

CEFR Common European framework of reference for languages

SFL Systemic Functional Linguistics

RST Rhetorical Structural Theory

GRI Get ready for IELTS

LIST OF TABLES

Table 2.1 Conjunctions in Halliday and Hasan (1976), Halliday and Matthiessen

(2014), and the present study 29

Table 2.2 Lexical relations in Halliday and Hasan (1976), Tanskanen (2006), Martin (1992), Halliday and Matthiessen (2014), and the present study 30

Table 3.1 Assignment 36

Table 3.2 Conjunctions for elaboration 42

Table 3.3 Coding scheme for elaborating conjunctions 43

Table 3.4 Coding scheme for lexical relations 47

Table 4.1 Frequencies of elaboration across levels 51

Table 4.2 Effective and ineffective elaborations 52

Table 4.3 Frequencies of conjunctions for elaboration 53

Table 4.4 Conjunctions for effective and ineffective elaboration 53

Table 4.5 Lexical relations for elaboration 55

LIST OF FIGURES

Figure 3.1 Procedure of data analysis 38

ABSTRACT

Cohesion has been widely concerned in English as a foreign language (EFL) writing Motivated by Halliday and Matthiessen‘s elaboration (2014), this study focused on conjunctions and lexical relations for elaborating relations in EFL expository writing Conjunctions, in this study, were examined for each category of elaboration, namely, exposition, exemplification, and clarification Lexical relations consist of repetitions, synonyms, antonyms, and inclusion

Based on Halliday and Matthiessen‘s model of elaboration (2014), the analysis was conducted on 35 expository paragraphs written by intermediate-level learners at a suburban English center in Ho Chi Minh City The results of analysis indicated the extent of elaboration and its corresponding conjunctions and lexical relations in the EFL learners‘ expository texts The results revealed that the learners‘ texts demonstrated limited elaboration both in terms of quantity and quality Elaboration was not only inadequate but also involved problematic use of conjunctions and lexical relations The learners used a variety of elaborating conjunctions, but these were not reliable evidence for elaboration Likewise, a range of lexical relations were present yet largely ineffective for elaboration

The results contribute to understanding of how the EFL learners used conjunctions and lexical relations as resources for realizing elaboration in expository texts, rather than as isolated ties The findings have important implications for teaching writing including teaching conjunctions, pre-teaching vocabulary, and revision with respect to elaboration, as well as further research on the topic

CHAPTER 1 INTRODUCTION

1.1 BACKGROUND TO THE STUDY

In order to produce effective communicative text, it is important for EFL learners to attend to cohesion According to Halliday and Hasan (1976), ―the concept of cohesion

is a semantic one, which refers to relation of meaning that exist within a text‖ (p 4) The phenomenon of cohesion occurs when the interpretation of an element depends

on another in text In practice, cohesion is a criterion for assessing the writing skills in high-stake English proficiency tests such as IELTS and TOEFL Under the Bachman‘s model of communicative competence (1990, p 87), cohesion has an important part in textual competence of language users.However, the role of cohesion

in EFL writing has not reached unanimous consensus among researchers In general, there have been two major approaches to research on cohesion: one focused on cohesion as individual ties (a single instance of cohesion) in isolation; the other attempted to align cohesion with coherence These two approaches have underpinned the research on cohesion in EFL writing, manifesting in two opposing research strands

In the first strand, cohesion was examined as composed of individual ties For example, Khalil (1990) investigated frequencies of cohesive devices used in EFL written texts Zhang (2000) researched the number of cohesive devices in expository writing by EFL university students Similarly, Tran (2011) explored the frequencies

of cohesive ties in argumentative essays by EFL learners More recently, Crossley, Kyle, and McNamara (2016) examined cohesive devices at global, local, and text levels used by university students While cohesion has been examined from a range of research settings, it tended to be broken down into individual ties in isolation Many scholars (e.g., Crossley et al., 2016; Flowerdew, 2013) agree that cohesive devices do not operate in isolation

As a result, cohesion in this strand was often found distinct from coherence Connor (1984) pointed out that cohesion was poorly correlated with coherence in English learners‘ written texts Likewise, Khalil (1989) found similar findings when correlating cohesion with coherence in EFL writing Leki, Cumming, and Silva (2009) commented that cohesion does not do their job of tying in EFL texts, and these texts are uninterpretable despite presence of cohesive device Crowhurst (1980) thus recommend that the most informative research on cohesion in students' writing does not come from merely counting instances; instead, researchers should focus on differences that exist in the way a particular cohesive item is used

Meanwhile, the second strand attempted to intertwine cohesion together with coherence Morley (2006) analyzed the use of lexical cohesion as resources for structuring argument in English newspaper discourse Tanskanen (2006) investigated the use of lexical cohesion as contributing to coherence across different text types Alarcon (2013) examined the relationship between lexical cohesion and writing quality in argumentative essays written by undergraduate students in a Filipino university Berzlánovich and Redeker (2012) investigated the alignment between lexical cohesion and rhetorical structure in expository and persuasive texts

From this body of research emerges two important considerations for the present study First, the studies claimed to investigate relationship between cohesion and coherence While cohesion tended to be fixed since it was mainly based on Halliday and Hasan‘ classical model (1976), coherence has been approached from different perspectives For example, Connor (1984) used multiple measures (topical structural analysis and problem-solving model) to approach coherence of argumentative writing Meanwhile, Khalil (1990) adopted Grice‘s maxims of cooperation to measure coherence Berzlánovich and Redeker (2012) employed Rhetorical Structural Theory (RST) scheme of relations to examine coherence Although numerous studies have been conducted to investigate cohesion, research on cohesive resources for realizing rhetorical function has remained rather scarce

Second, while many studies were conducted on cohesion as realizing rhetorical functions, they were mainly based on native speakers‘ data rather than EFL written texts Morley (2006) was interested in newspaper discourse written by professional writers Similarly, Berzlánovich and Redeker (2012) investigated lexical cohesion aligned with rhetorical structure of texts by native language users The only exception that could be found is Alarcon (2013), who was concerned with rhetorical functions of lexical cohesion in argumentative writing by EFL learners in the Philippines There is, then, a lack of research into cohesion that is aligned with rhetorical in the context of EFL learner writers, especially at English centers Thus, it is still unclear whether the use of cohesive devices contribute to rhetorical functions in EFL written texts

The present study finds itself in the second research strand on cohesion, that is,

as contributing to rhetorical organization of texts Its focus is on expository writing as this genre is required by the research setting, on which the present research is based Moreover, when comparing (lexical) cohesion between expository texts and persuasive texts, Berzlánovich and Redeker (2012) found that lexical cohesion was more closely aligned with rhetorical structure of expository texts than that of persuasive texts Informed by these researchers, expository writing provides a suitable context for investigating cohesion However, unlike Berzlánovich and Redeker, this present study only concerns elaboration, a rhetorical function that is required much in this genre (Zhang, 2000)

1.2 AIM OF THE STUDY

Focusing on elaboration, the present study places itself in the Halliday and Matthiessen‘s model of elaboration (2014) as it not only addresses cohesive relations but also aligns them with elaborating relations, which corresponds to rhetorical functions featured in expository writing The discussion on the framework and its relevance to expository writing will be provided in Chapter 2 For now, motivated by the model, the study aims to explore the use of conjunctive and lexical relations for elaboration in expository paragraphs by intermediate-level EFL learners

1.3 RESEARCH QUESTIONS

Informed by the research aim, it is now possible to formulate the main research questions concerning conjunctive and lexical relations aligned with the rhetorical functions of expository writing for the present study:

1 What conjunctive relations are used for elaboration in expository paragraphs by intermediate-level EFL learners?

2 What lexical relations are used for elaboration in expository paragraphs by intermediate-level EFL learners?

1.4 SCOPE OF THE STUDY

The present investigation concerns intermediate-level EFL learners, who have little formal training in English linguistics, especially on lexical relations The knowledge they gained on conjunctions and lexical relations came from the instructions in the

textbook Get Ready for IELTS (Aish, Short, Snelling, Tomlinson, & Geyte, 2016)

However, this background does not mean that the learners are incapable of using conjunctions and lexical relations Just as a language user does not need formal training in linguistics to to use lexical relations, the EFL learners can use the linguistic resources in their writing, either consciously or unconsciously

The present study is confined to the expository paragraph only Though this scope is quite narrow, the level of paragraph plays fundamental role in the process of learning of the participants It lays a building block for transitioning to writing essays Moreover, paragraph makes a suitable choice for the analysis of elaboration and its corresponding lexical relations These are originally relationships between clauses (Halliday & Matthiessen, 2014), so applying them to the level of paragraph, i.e., topic sentence and supporting sentences is more viable than some higher units like essays

According to Halliday and Matthiessen (2014), elaboration is a set of semantic relationships However, the author notes that it is important to interpret these

logical-‗logical‘ relationships in their own terms as part of the semantics of a language, and not to expect them to fit exactly into formal logical categories – although since the

latter were derived from natural language in the first place there will obviously be a close relationship between the two (p 460)

1.5 SIGNIFICANCE OF THE STUDY

One of the starting points for this project is the assumption that cohesive devices are not used effectively to tie EFL texts as informed by previous studies (Connor, 1984; Khalil; 1990; Zhang, 2000), including expository ones For this reason, the present study does not aim to challenge these findings; rather, it describes the use of cohesive relations as contributing to elaboration of expository paragraphs The outcome could

be utilized to show learners where their use of cohesive devices is effective or ineffective, not in isolation but for elaboration In its turn, although elaboration is only basic moves in the whole system of rhetorical structure in text, it is much used in expository writing (Zhang, 2000)

1.6 ORGANIZATION OF THE THESIS

This thesis consists of six chapters Following this introduction, Chapter 2 reviews relevant literature on expository writing to supply a clearer context for addressing the research questions It considers Halliday and Matthiessen‘s model elaboration (2014) for understanding the concepts of elaboration, conjunctions, and lexical relations The chapter also pay attention to conjunctions and lexical relations as constrained by elaboration Chapter 3 explains the methods that have been used for analyzing data in this study It provides details of the research site and participants, as well as the procedures used for data collection and analysis Chapter 4 presents the results of analysis aimed at answering the two research questions These results are then discussed in relation to earlier studies on conjunctions and lexical relations Chapter 5 summarizes major research findings as well as the limitations, based on which it points out implications for teaching practice and further research on the topic of elaboration in EFL writing

CHAPTER 2 LITERATURE REVIEW

The first chapter draws attention to the motivations for investigating conjunctive and lexical relations as resources for elaboration in EFL learners‘ expository texts This chapter searches for an analytical model to address these concepts First, it starts with

a discussion on the expository genre to determine the context within which elaboration that suits this research context of EFL Second, based on the characteristics of expository genre, relevant models for elaboration as well as conjunctions and lexical relations are discussed in detail Finally, the chapter concludes by proposing an adapted model of analyzing conjunctive and lexical relations for elaboration in expository paragraphs

2.1 EXPOSITORY WRITING

This present study essentially deals with the text-forming relations, i.e., elaboration, conjunctions, and lexical relations Moreover, these relations are not general but particular to expository paragraphs For these reasons, it is important to determine what text-forming relations relate in the expository writing This aim is realized by reviewing the notion of expository genre, its features at the level of paragraph

The term genre refers to a type of discourse in a particular setting that has

distinctive and recognizable organization and structure, and that has particular and distinctive communicative functions (Richards & Schmidt, 2010, p 245) Examples

of genre include business reports, news broadcasts, letters, advertisements, etc From the Systemic Functional Linguistics (SFL) perspective, Martin (1992) defines genre

as ―a staged, goal-oriented social process realized through register‖ (p 505) Hyland (2009) explains that ―social processes because members of a culture interact to achieve them; goal-oriented because they have evolved to achieve things; and staged because meanings are made in steps, and it usually takes writers more than one step to reach their goals‖ (p 63) Hyland highlights that the SFL-oriented genre is significant

in the field of language teaching

Biber (1989) proposes three expository categories on a linguistic basis, including scientific exposition, learned exposition, and general narrative exposition The author explained that all of these are expository since they feature highly informational density and noninvolved production However, general narrative exposition differs from the others in that it uses narration to convey information In addition, the narrative amount is neither imaginary nor entertaining; rather, it is an integral part of the expository information being conveyed Among the three types, general narrative exposition seems most relevant to the present study, in which writing tasks are general rather than technical or specific to a particular domain of subject matter, i.e., engineering or biology However, the study deals with cohesive relations between sentences, while Biber mainly focused on linguistic features at the sentence level Thus, Biber‘s notion of general narrative exposition is not compatible for purpose of the current study

Reid‘s (1982) provides an account of exposition that seems more compatible

with Martin‘s definition of genre In terms of goal, exposition is used for explanation,

as distinguished from persuasion (Reid, 1982, p 31) Staged may reflect in that the

main goal of exposition is achieved in several steps Reid added that the goal of exposition is to present a viewpoint to the reader, which is then to be explained,

clarified, and illustrated in the sub-goals The social aspect manifests in that

exposition is written to present a viewpoint to an audience (Reid, 1982, p 99)

These features embody in Reid‘s account of exposition at the paragraph level

The author mentions the paragraph consists of a topic sentence and a series of

supporting sentences Topic sentence is the most general and important sentence in an academic paragraph (Reid, 1982, p 11), directing the focus of the paragraph that follows The sentence has a controlling idea that the following sentences will function

on, i.e., to explain, define, clarify, and illustrate Reid (1982, p 12) further explains that ―a controlling idea is a word or phrase that can be asked questions about,‖ for example, how, why, in what ways or what it means

These paragraph features, though proposed a long time ago, still exert influence

on EFL writing instructions, especially the present setting, whose paragraphs have following characteristics:

 A paragraph should have one main idea

 A paragraph should have more than one supporting idea

 The first sentence of a paragraph should contain the main idea of the

paragraph

 The last sentence of a paragraph should contain the main idea of the

paragraph

(Aish et al., 2016, p 13) Reid‘s account could be criticized for being too rigid The notion of topic sentence lends itself to the writer-responsible tradition, which means that it is the writer who is responsible for producing a text that will be coherent to the potential reader (Celce-Murcia & Oshtain, 2000, p 149) According to Allison, Varghese, and

Wu (1999), such a view of writing seems to be an imposition rather than a negotiation

of meaning and discourse development However, despite holding reservations about topic sentence, the researchers agreed it is useful to help learner writers to become more aware of reader expectations for pedagogy Indeed, Reid (1996) argued that the account of expository paragraphs is highly beneficial for inexperienced EFL writers in learning academic writing and using topic sentences can help the learners to communicate effectively and efficiently

There have been numerous studies on cohesion relevant to EFL expository writing For instance, Zhang (2000) analyzed cohesive features in the expository writing of Chinese university students Adiantika (2015) investigated cohesive devices that EFL students used in expository writing and their contribution to the written products These studies had in common the neglect of features of expository paragraphs, including the relationship between the topic sentence and supporting sentences, and rhetorical functions the sentences serve Alarcon (2013) took a further

step by attending to rhetorical functions of lexical relations; however, the considerations seemed more general than specific to the genre under scrutiny

Despite investigating cohesion in the expository genre, these studies largely neglected the features of expository paragraphs described so far in this section, whether its communicative functions or hierarchical unity They focused merely on how frequently cohesive devices were used, hardly addressing the relation of cohesion

in the context of exposition Tanskanen (2006) analyzed the role of lexical cohesive relations in expository texts, showing that combining sentences which had an average

or an above average number of cohesive units results in an intelligible summary As such, although Tanskanen dealt with expository texts, it had a very different purpose from this present study; moreover, her study on lexical relations was based on native speakers‘ data rather than EFL learners Likewise, Berzlanovich and Redeker (2012) were concerned with lexical relations aligned with the rhetorical structure of the expository genre The researchers found that the lexical resources are valuable clues

to its organization However, since these focused on native speakers‘ data, it is not certain whether their findings apply to EFL writing or not

In conclusion, the text-forming relations in this study are not general but specific

to a particular text type, that is, expository paragraphs Based on Reid‘s account, the relations necessarily relate the topic sentence with supporting sentences Despite being proposed a long time ago, Reid‘s account of expository paragraph turns out to

be feasible for the purpose of this present study The features of expository paragraphs serve to specify the context in which elaboration as well as conjunctions and lexical relations are examined

2.2 HALLIDAY AND MATTHIESSEN’S MODEL OF ELABORATION

2.2.1 Terminology

The term elaboration can be found in several studies on cohesion and coherence For

example, Khalil (1990) attributed elaboration to informativeness, which was more relevant to content than cohesion On the other hand, elaboration is a subset of

functional relations in Rhetorical Structure Theory (RST) developed by Mann and Thomson (1988) Rhetorical structure is the underlying structure which accounts for the organization of a text or discourse (Richards & Schmidt, 2010) Different kinds of texts and discourse (e.g., stories, descriptions, letters, reports) are distinguished by the ways in which the topic, propositions, and other information are linked together to form a unit

RST elaboration is a type of functional relations between two parts of text The part that plays a dominant role is called nucleus; the other that plays a secondary role

is called a satellite The satellite presents added details about the situation, or some element of subject matter presented in the nucleus or inferentially accessible in the nucleus The relatedness is recognizable in one or more of the ways listed below:

(Mann & Thompson, 1988)

In the list, if the nucleus presents the first member of any pair, then the satellite includes the second By virtue of RST elaboration, the reader may recognize the situation presented in the satellite as providing additional detail for the nucleus The reader also identifies the element of subject matter for which detail is provided

As functional relations in texts, RST elaboration does not rely on linguistic realization for their identification (Mann & Thomson, 1988; Taboada & Mann, 2006)

As such, a problem with RST is that its method of textual analysis is inevitably subjective Taboada and Mann (2006) insist that certain relations are ―rarely or never signaled‖ Admittedly, the language system does not have overt signals for realizing

all relations; there exist relations which are not identifiable by specific linguistic means Nevertheless, this reason does not qualify for this present study, which concerns elaboration only, not all possible relations Moreover, without bearing on linguistic realizations, benefits may lose from the perspective of language education Hoey (1983, cited by Flowerdew, 2013) states that ‗discussion of types of relation cannot be sensibly carried on apart from the means whereby those relations are identified (p 51)

Elaboration in this present study comes from expansion, a broader set of

logico-semantic relations modeled by Halliday and Matthiessen (2014) Expansion involves the secondary clause expanding the primary clause, by means of elaboration, extension, or enhancement (Halliday & Matthiessen, 2014, Chapter 7) As such,

besides elaboration, expansion includes two other subsets of relations: extension and

enhancement In extension, one clause expands another by extending beyond it,

adding some new element, giving an exception, or providing an alternative Enhancement involves one clause qualifying another with some circumstantial feature

of time, place, cause, or condition Elaboration occurs when ―one clause expands another by elaborating on it (or some part of it): restating in other words, specifying in greater detail, commenting, or exemplifying‖ (Halliday & Matthiessen, 2014, p 444) Out of the three types of expanding relations, elaboration is particularly relevant

to the current study As discussed in 2.1, the expository paragraph features hierarchical unity, which reflects in the relation of the topic sentence and its supporting sentence This unity seems to be more compatible with elaboration than extension or enhancement In the context of expository paragraphs, if a sentence extends or enhances the topic sentence, it seems less likely to maintain the most general and important ideas Extending relationships would entail more digression than unity Similarly, enhancement serves to qualify the central idea with additional elements instead of elaborating them These do not mean that extension and enhancement never occur in expository writing; they are not directly contributing to the unity of expository paragraphs, hence outside the scope of the study Meanwhile,

with elaborating relationships applied to the topic sentence and its supporting sentences, the main idea and the unity of the expository paragraph would be maintained and tracked

Elaboration from RST and from Halliday and Matthiessen‘s model (2014) shares one thing in common Mann and Thompson (1988) assert that the meaning of

RST elaboration deals with subject matter relations, which are for readers to recognize the relation under scrutiny Subject matter relations are distinct from

presentational relations, which aim to increase some inclination in the reader, such as

the desire to act or the degree of positive regard for, belief in, or acceptance of the dominant text part

Likewise, Halliday and Matthiessen‘s elaboration operates on experiential

meaning, which corresponds to subject matter This could be traced back to the notion

of logico-semantic relations Under the SFL systems, these logico-semantic relations belong to the ideational function of language, which is divided into experiential and logical meaning (Martin, 1992) Accordingly, the relations relate sequences of figures (or moves) that are presented as textually related messages (Halliday & Matthiessen,

2014, p 428) The experiential meaning expresses itself through the notion of thesis,

which Halliday and Hassan (1976) refers to as the content of what being said, coinciding with experiential events coded in the text (p 239)

However, elaboration of Halliday and Matthiessen (2014) differs from that of RST in one important aspect RST elaboration is not tied to linguistic signals while Halliday and Matthiessen‘s elaborating relations are closely associated with linguistic resources for cohesion Elaborating relations are accompanied by conjunctions and lexical relations In this virtue, Halliday and Matthiessen‘s elaboration creates conditions for investigating the cohesive resources for the rhetorical relations in expository paragraphs

2.2.2 Categories of elaboration

In Halliday and Matthiessen (2014), logico-semantic relations interact with the system

of taxis, based on which elaboration is categorized into hypotactic and paratactic Hypotactic elaboration is a way of ―introducing into the discourse background information, a characterization, an interpretation of some aspect of the dominant clause, or some form of evaluation‖ (p 464) Hypotactic elaboration involves a special clausal construction – that of the non-defining relative clause, or finite clauses Although hypotactic elaboration may well relate two clauses, its boundary is limited

to sentence rather than across the text At the same time, paratactic elaboration involves an ordinary clause without any special grammatical constraints Thus, the present study is more concerned with paratactic elaboration

Paratactic elaboration involves three further subcategories: exposition,

exemplification, and clarification Exposition at this stage needs not to be confused

with exposition as a genre expounded in 2.1 Elaborating exposition occurs when one

clause restating the thesis of another in different words, to present it from another

point of view This often involves repetition or synonymy For example, in ―Yeah, I

just hate it; I just loathe it‖ (Halliday & Matthiessen, 2014, p 463), the second clause exposes the first by the use of synonym loathe for hate Exposition finds itself

equivalent to RST restatement, a functional relation proposed by Mann and

Thompson (1988) RST restatement involves a text segment restating another This relation has the effect that reader recognizes one sentence as a restatement of the other The authors provide the example:

Well-groomed car reflects its owner The car you drive says a lot about you Between the two sentences is a relation of restatement, in which the content in the first sentence is restated in the second one

Exemplification occurs when one clause develops the thesis of another by becoming more specific about it, often citing an actual example For example, in ―We

used to have races– we used to have relays” (Halliday & Matthiessen, 2014, p 463), the latter clause serves as an example for the former with relays being a hyponym of

races In clarification, one clause clarifies the thesis of another, backing it up with

some form of explanation or explanatory comment Clarification may also be an evaluative comment or involves a shift in positive to negative For example, ―I wasn‘t

surprised, – it was what I had expected‖ (Halliday & Matthiessen, 2014, p 464), the

latter clarifies the former with the use of expected as an antonym of surprised

In Halliday and Matthiessen‘s model of elaboration, exposition is accompanied

by repetition, synonymy and exemplification is tied with hyponymy and meronymy While exposition and exemplification are associated with specific categories of lexical relations, clarification is not explicitly indicated in this regard The absence of lexical relations may indicate variations in how a clause can be clarified However, based on the authors‘ account on clarification and its examples, clues to antonymy can be recovered Clarification often involves a shift of polarity from positive to negative or

vice versa In the example ‗they weren‘t show animals; we just had them as pets‘ (p

464), it is reasonable that the type of lexical relations most relevant to the shift of polarity is antonymy; for this reason, this study adds antonymy to the framework for clarification to make up for the lack of explicit involvement of antonymy in the original framework Overall, lexical relations involved in elaboration include repetition, synonymy, hyponymy, meronomy, and antonymy

Elaboration in the present study is based on Halliday and Matthiessen‘s model

of elaboration When translated to the context of expository paragraphs, the relations may have a limited scope of effect, corresponding to restatement, clarifications, and examples However, by means of these functions, elaboration has the potential to maintain the textual unity as required by expository paragraphs Moreover, elaboration accounted by Halliday and Matthiessen has the advantage of being aligned to cohesive resources, conjunctions, and lexical relations

2.3 CONJUNCTIVE RELATIONS FOR ELABORATION

In the classical model of cohesion (Halliday & Hasan, 1976), conjunctive and lexical relations were kept as independent categories of cohesion By contrast, under the framework of elaboration, the cohesive resources are closely aligned with each

category of elaboration This section places the conjunctions back on its developments

to determine their meaning and theoretical underpinnings Also, it surveys how the conjunctions have been investigated in EFL texts, especially for the expository genre

2.3.1 Terminology

In the present study, conjunctions belong to the model of elaboration The term

conjunction has also been known in different names, for example, connectives

(Crossley et al., 2016), connectors (Granger & Tyson, 1999), and transition signals (Oshima & Hogue, 2007) Familiar as they are in writing instruction, these terms seem inconsistent with the system of cohesion developed by Halliday and Hasan (1976), hence not as preferable as conjunctions This section aims to expose characteristics of conjunctions by comparing Halliday and Hasan‘s classical model (1976, Chapter 5) and Halliday and Matthiessen‘s model of elaboration (2014)

Overall, conjunctions have been approached in two opposing ways On the one hand, the meaning was considered as contained in itself (Christiansen, 2011, as cited

in Flowerdew, 2013, p 38) This could be observed in studies that counted the frequencies of conjunctions Khalil (1990) investigated how frequently conjunctions EFL learners used as part of cohesion in written texts Similarly, Zhang (2000) counted the frequency of conjunctions as a measure for general cohesion As a result,

it was likely that the conjunctions were treated in isolation without influence from textual environment On the other hand, conjunctions were viewed as relations of meaning between entities, which was based on the presence of conjunctive elements, (Taboada & Mann, 2006) Concha and Paratore (2011) adopted conjunctions as a means to investigate coherence relations in persuasive texts by Chilean students

Under the classical system of cohesion, conjunctions differ from other cohesive resources in that they specify the way in which ―what is to follow is systematically connected to what has gone before‖ (Halliday & Hasan, 1976, p 277) Accordingly, conjunction is a relation of meanings, manifesting in two regards First, the relation between meanings reflects in the content, language users‘ experience of external

reality, hence called experiential or external meaning (Halliday & Hasan, 1976, p

240) Second, the relation between meanings manifest itself in the sense that stands for the speaker‘s choice of speech role and rhetorical channel, his attitudes, his

judgement and the like, thus called internal meaning

This internal/external distinction has been rarely addressed in studies on conjunctions in EFL writing For example, Johnson (1992) considered conjunction as cohesive without discerning their internal and external meanings, i.e., by its semantic relation between two events, two stages in an argument, or two continuous passages

of a text The only exception that could be found is a study by Field and Lee (1992), who compared the use of internal conjunctions in texts by EFL and native speakers For the researchers, internal conjunctions occur when the writer uses a device to make

a conscious juxtaposition of one point to another Martin (1992) warned that many of the same conjunctions can realize both internal and external relations Thus, problems may arise in deciding whether an internal or external interpretation is appropriate Georgakopoulou and Goutsos (2004, as cited in Flowerdew, 2013) argued that the strongest meaning of discourse markers is not ideational, but interpersonal For

instance, y’know at the beginning of an utterance appears to indicate attitudinal

meaning to what is being said

Dealing with the internal-external distinction of meaning, Martin (1992, p 226) suggests that a useful test for deciding on the proper meaning is to change the dependency relationship between the messages under scrutiny (by paraphrasing from hypotactic to paratactic or "cohesive" or vice versa) However, the author notices that despite the usefulness of the paraphrase test, it may be difficult to make the internal and external distinction of, for example, additive relations The reason is with these relations, the difference between constructing text and constructing field is not always clear-cut

The present study does not concern conjunctions in general but for elaboration, a set of logico-semantic relations, which pertain to experiential and logical meaning

(Martin, 1992) In line with elaboration, conjunctions are more concerned with experiential meaning than interpersonal meaning The experiential meaning of

conjunctive relations manifests in the notion of thesis, the content of the event that is

being said (Halliday & Hasan, 1976, p 239)

In conclusion, the classical conjunctions and elaborating conjunctions differ with respect to internal and external meaning In the classical version, meaning of conjunctions is mainly internal, though external meaning is possible In the elaborating conjunctions, external meaning seems more significant than internal meaning, which results from conscious juxtaposition of ideas (Field & Lee, 1992) Compared with the classical conjunctions, then, the expository and exemplifying conjunctions are no longer limited to internal communication process but linked to external reality as well

2.3.2 Categories of conjunctive relations

Originally, conjunctions under Halliday and Hasan‘s classical scheme (1976, p 238) include four categories: additive, adversative, causal, and temporal Meanwhile, elaborating conjunctions involve three categories: expository, clarification, and exemplification, whose origin is traced back to additive and adversative conjunctions

in the classical scheme These two categories of conjunctions are discussed in more detail with comparison between the classical model and the elaborating scheme The classical additive relations (Halliday & Hasan, 1976) cover four sub-groups: simple, complex, comparative, and appositive Simple additive relations carry both external and internal meanings Typical expressions for simple additive relations

include and, and or The relations can also be expressed negatively by, for example,

nor, and neither Complex additive relations belong mainly to the internal plane The

emphatic forms include furthermore, moreover, and besides that; The de-emphatic forms include incidentally, or by the way Comparative additive relations are mainly associated with internal meaning, which can be expressed by similarly, likewise, and

by contrast The three categories are very common in text but are not focused on in

the present study Instead, among the four categories, it is the appositive relations are

the most relevant to elaborating conjunctions In the 1976 model, appositive relations are resources for realizing both internal meaning and external meaning Common

expressions for these relations include in other words, to put it another way, and for

example

Adversative conjunctions basically mean ‗contrary to expectation‘ (Halliday & Hasan, 1976, p 250) The expectations may be derived from the content of what is being said or from the communication process, which means that the cohesion can be found to carry both external meaning and internal meaning In the internal sense, expectation does not come from what a presupposed sentence is about; instead, the point reached in the communication process The meaning is ‗in spite of the roles the speakers are playing, or the state of the argument‘ Another sense of internal

adversative relations is ‗as against‘, which is expressed by items such as in fact, as a

matter of fact, and actually The meaning is something like ‗as against what the

current state of communication process would lead us to expect, the fact of the matter

is …‘ The conjunctions take the form of an assertion of truth (Halliday & Hasan,

1976, p 253)

Under the classical scheme, adversative relations include four categories: proper, dismissive, contrastive, and corrective Proper adversative relations encompass both internal and external meanings, which can be realized by common expressions such as

yet, though, however, etc Dismissive adversative relations have the meaning where

some circumstances are dismissed as irrelevant (p 254), which may be both external

and internal Some common expressions for these relations include whichever

happens, and whether … or not Similarly, contrastive relations belong to both

external and internal planes The external meaning of these relations can be expressed

by for example but, however, and at the same time The internal contrastive relations, which are also called avowal, mean ‗as against what the current state of the

communication process would lead us to expect, the fact of the matter is …‘ (p 253)

Avowal is realized by such expressions as in fact, actually, or as a matter of fact

Corrective relations are closely related to the avowal and also belong to the internal

plane These relations can be signaled by instead, rather, or on the contrary

The original conjunctions have been investigated in numerous studies on cohesion in EFL writing Zhang (2000) investigated the number of conjunctions in expository texts by EFL university students The researcher reported that conjunction category accounted for 17.5% of the cohesive devices Zhang correlated frequencies

of the conjunctions and the learners‘ writing quality Likewise, Tran (2011) adopted the same approach to the classical conjunctions and correlated conjunctions with writing quality Adiantika (2015) investigated the contribution of conjunctions to expository writing quality The researcher found that conjunctions occurred 181 times, the second most common type of cohesion in expository writing

Generally, these studies treated conjunctions as isolated items rather than a relation of meaning as Halliday and Hasan (1976) emphasized in the first place The conjunctive ties were often counted individually as a measure for cohesion For example, Johnson (1992) calculated that conjunctive ties under the syntactic ties rather than for their own right Zhang (2000) found that conjunctive ties in EFL texts were less common than lexical ties and more common than reference ties Among conjunctions, additive devices were the most common, followed by temporal, causal, adversative, and continuative (Liu & Braine, 2005) Approached as mere numbers, the conjunctions were no more than surface links whose meaning was contained in themselves, as if they existed independently from textual environment This approach made conjunctions analogous with the notion of text as an autonomous object, which can be described independently of context (Hyland, 2009, p 8)

As motivated by the elaborating framework and expository paragraphs, this study reserves the status of conjunctions as relations of meaning, as proposed by Halliday and Hasan (1976), not merely surface markers It follows that on the one hand, the conjunctions are based on conjunctive elements; on the other hand, they do not exist in themselves but closely related to a specific category of elaboration Such

expressions as for example, for instance, and in particular may accompany the exemplifying elaboration For exposition, typical conjunctions include or (rather), in

other words, that is to say or I mean; or, i.e Likewise, expressions such as in fact, actually, indeed, at least are commonly used for clarifying elaboration These

conjunctions, thus, do not exist alone but are lined up with elaboration between a sentence and its precedent

In conclusion, the elaborating conjunctions have their origin as part of additive and adversative relations: expository, exemplificatory, avowal, and corrective relations The account shed light on three issues bearing upon on how conjunctions could be analyzed: internal or external meaning, typical realizations, and textual environment Although the scope of elaborating conjunctions is rather narrow compared with the classical model, they are more suitable for the purpose of this study in that they are lined up with elaborating functions, i.e., exposition, exemplification, and clarification This characteristic creates conditions for evaluating how the conjunctions are used in EFL texts, whose cohesion is considered as not tying texts (Leki, et al., 2009)

2.4 LEXICAL RELATIONS FOR ELABORATION

Compared with conjunction, the lexical cohesive resources are reserved more space of discussion in this review chapter Lexical cohesive relations are particularly problematic (Tanskanen, 2006; Flowerdew, 2013) This section describes the nature

of lexical relatedness, and the role of context, whereby to inform how they are analyzed in the present study

2.4.1 Terminology

In general, the term lexical relation refers to semantic relations among word such as

antonymy, synonymy, and hyponymy (Murphy, 2003) One alternative term adopted

by semanticists is sense relation (Hurford, Heasley, & Smith, 2007, p 26) According

to the authors, sense refers to relationships that are internal to the language system

Adopting sense relation plausibly entails exclusion of reference, which refers to

relationships between language and the world Indeed, Murphy (2003, p 11) argues that senses are not the only determining factor of lexical semantic relations Sense

relation sound too narrow given the scope of the present study, which concerns lexical

relations as cohesive devices in texts

Lexical relations in this study are a component of elaborating model described in 2.2 The term lexical relation is more systematically consistent with their origin from lexical cohesion (Halliday & Hasan, 1976) As part of cohesion, lexical relations are relations of meaning in texts Also, dealing with meaning, lexical relations are inclusive of both sense and reference rather than limited to either

2.4.2 Categories of lexical relations

Lexical relations are not only resources for cohesion but also for realizing elaboration Based on these constraints, this section searches for a taxonomy of lexical relations that best serve the purpose of the present study

In their classical work, Halliday and Hasan (1976) divided lexical cohesion into

two categories: reiteration and collocation Reiteration involves repeating of a lexical

item, or using a general word, synonym, a near-synonym, a superordinate (Halliday & Hasan, 1976, p 278) Collocation, on the other hand, covers both systematic lexical relations and even unsystematic relations Systematic collocational relations include pairs of opposites like complementaries, antonyms, and converses, for example

boys—girls, and stand up—sit down The group of opposites are from now on referred

to as antonymy for better consistency with the terminology of elaboration Besides antonymy, collocational relations include pairs that stand in part to whole, and part to part relations, both of which are also called meronymy following Halliday and Matthiessen (2014) In addition, collocation also includes pairs that belong to hyponymy, he relations between groups of words all falling under one superordinate

For example, apple, orange, banana, and lemon are all hyponyms of the superordinate

fruit Besides systematic relations, Halliday and Hasan included under collocation

relations that were difficult to describe in a systematic way, for example, laugh—joke,

blade—sharp, garden—dig

Based on the model of Halliday and Hasan, numerous studies found that lexical relations were the most important cohesive resources Khalil (1990) reported that lexical relations accounted for the largest proportion of cohesive devices in EFL texts Liu and Braine (2005) found that lexical devices formed the largest percentage of the total number of cohesive devices, followed by references and conjunctives Similar findings were also obtained by Adiantika (2015) with lexical cohesion being the most common devices with 322 occurrences in nine EFL writings

However, these studies revealed that the use of lexical relations was rather limited in EFL writing For example, when comparing the categories of lexical relations with one another, Liu and Braine (2005) found that in the five sub-categories

of lexical devices used by EFL learners, repetition accounted for the highest percentage, followed by collocations, synonyms, and antonyms The researchers concluded that EFL learners demonstrated a limited choice in the use of lexical items and the great majority of the lexical devices were repetitiously used Similarly, Alarcon (2013) investigated different lexical relations as resources for cohesion in EFL argumentative texts This study also found repetition with 32.42% was the most frequently used lexical cohesive device, followed by synonyms with 31.64% Among the types of lexical relations, hyponyms and meronyms were the least frequently used, occupied 11.68% and 7,69% respectively

The classical model of lexical cohesion was considered by Stotsky (1983) as lacking a consistent principle for classification The author then proposed reorganization of lexical cohesion based on systematicity of lexical relations This revised model included all recognizable systematic relations under reiteration, while reserving unsystematic relations under collocation This reorganization later found it similar to Tanskanen‘s model of lexical cohesion (2006) Since the two models were based on the same principle of organization, this section pays more attention to the newer model of Tanskanen

Tanskanen (2006) aimed at analyzing lexical cohesion as contributing to coherence in texts of different types, including expository writing Like Halliday and

Hasan‘s (1976), this model also consisted of two categories: reiteration and

collocation However, while Halliday and Hasan‘s reiteration includes only repetition,

synonym, superordinate, and general words, Tanskanen‘s reiteration expanded to include systematic semantic relations such as synonymy, hyponymy and antonymy Tanskanen also included under lexical reiteration the devices of grammatical cohesion, namely, pronouns As a result, Tanskanen‘s reiteration was divided into eight subcategories: simple repetition, complex substitution, equivalence, generalization, specification, and contrast (Tanskanen, 2006, p 61) Meanwhile, collocation under Tanskanen‘s model consisted of unsystematic relations such as ordered sets, activity-related collocation, and elaborative collocation

Perhaps the most notable difference between Halliday and Hasan (1976) and

Tanskanen (2006) was terminology Instead of synonymy, Tanskanen adopted the

term equivalence While Halliday and Hasan used hyponymy, Tanskanen chose the term generalization As for antonymy, Tanskanen adopted contrast when referring to

the relation between an item and another item which has an opposite meaning Only meronymy carried a different interpretation from that used by Halliday and Hasan (1976) Instead of part-whole relationships, meronym referred to the relation between

an item and a more specific item In other words, meronymy is different from hyponymy in their opposing directions: hyponymy goes from specific to general while meronymy starts from general to specific

Tanskanen explained that the terminology was adopted to avoid linguistic terminology and account for specificity of discourse According to the researcher, the non-linguistic terms emphasize the fact that the justification and explanation for a relation between lexical items can and should be found in their textual environment

As such, besides terminology, the types of lexical relations proposed by Tanskanen are not so different from those of Halliday and Hasan (1976) Difference, if any, lies

in the way Tanskanen reorganized the relations into reiteration and collocation However, since the present study investigates the elaborating functions of lexical relations, it is interested in the relations themselves rather than under reiteration or collocation Therefore, instead of Tanskanen‘s new reorganization, this present study retains the lexical relations set out in Halliday and Hasan‘s original model

Another account of lexical relations that is closely related to Halliday and Hasan‘s model is the taxonomy described by Martin (1992) These lexical relations are organized under the two sub-systems: superordination and composition (Martin,

1992, p 295) Superordination reflects taxonomies based on subclassification Composition reflects taxonomies based on relation of parts to whole People, places, and things as well as actions are usually organized with respect to both types of taxonomy However, only superordination, not composition, can be used to organize qualities

Lexical relations under superordination include hyponymy, hyperonymy, and co-hyponymy Hyponymy refers to relations in which a subclass follows its class

(e.g., sax-alto) Hyperonymy applies to the converse sequence where the subclass appears first (e.g., alto-sax) Closely related to hyponymy is the category of

synonymy According to Martin, synonyms might well be defined as co-hyponyms for which differences in meaning do not matter The author relates the difference between synonyms and co-hyponyms to delicacy with respect to a particular field (the contextual projection of experiential meaning) The difference of synonyms and co-hyponyms also concerns experiential and interpersonal meaning Items that might be treated as synonyms experientially often differ with respect to the attitudinal meaning The difference between synonymy and co-hyponymy thus can be regarded as a kind

of scale moving from the same meaning to another meaning (Martin, 1992, p 301) Also related to the system is repetition, which the author considers as true synonymy, including derivative variations of a lexical item

Another category that is also related to co-hyponymy is antonymy As against

synonyms, antonyms are co-hyponyms for which differences in meaning do matter since items are opposed in meaning instead of complementing each other The

account of antonymy by Martin (1992) includes converse and contrast Converse

involves a variety of relations, including reciprocal social roles, kinship relations, and location in time and space Contrast refers to non-binary oppositions including two

types: series and cycles Series feature outmost members of a set (e.g., fail) while cycles just arrange items between two others (e.g., Monday-Tuesday-

credit-pass-Wednesday-Thursday-Friday-Saturday-Sunday) Antonyms and series may involve

gradable or non-gradable oppositions where they involve qualities of people, places, and things The gradability could be tested by whether an item can be intensified or

compared Lyons (cited by Martin, 1992) reserves the term antonym for gradable oppositions, introducing complementarity for gradable ones However, the relations

will all be referred to as antonymy for consistency with elaborating relations as mentioned in 2.2

Compared with superordination, composition is simpler with only one category This category organizes people, place, things, and actions based on relation of parts to whole The following items are among those realizing relational meronymy: part, content, ingredient, fitting, member, constituent, stratum, rank, plane, element, factor, component, faction, excerpt, extract, selection, piece, segment, section, portion, measure (Martin, 1992)

In conclusion, the taxonomy of lexical relations does not vary significantly across the three models Although the models have different underlying principles of categorization as well as slightly different names, they all include repetition, synonymy, hyponymy, meronymy, and antonymy Under the model of elaboration, lexical relations need to include not only those systematic recognizable relations Halliday and Hasan (1976, p 18) caution that cohesion is not only a general relation

in the language system but also a process in the text, in the sense that it is instantiation

of the general relation in real time; for this reason, relations should be described in

text or discourse specific manner As such, Martin‘s context-specific account demonstrates useful insights as to how each type of lexical relations can be identified

in texts The usefulness of Martin‘s taxonomy for elaborating relations is discussed in the next subsection

2.4.3 Lexical relations and context

From the discussion on categories of lexical relations has emerged underlying factors that determines how they are categorized: systematicity and field-specificity

As Stotsky (1983) pointed out, Halliday and Hasan (1976) lacked a principle in classifying lexical relations Tanskanen (2006) took up Stotsky‘s idea of reorganizing lexical relations surrounding systematicity However, the notion of systematicity is not relevant to analyzing lexical relations for elaborating relations While lexical relations are influenced by context, the way context was treated varied across the three models being discussed so far

In the seminal book Cohesion in English, Halliday and Hasan mainly mentioned

two types of contexts: situational and linguistic Context of culture generally means socio-cultural background, which was also mentioned but too far beyond the scope of cohesion Context of situation refers to the environment in which a text is taking place, whereby enables language users to accept a passage as text (Halliday & Hasan,

1976, p 18) This source of information needed for interpreting some elements in the text is not to be found in the text at all, but in the situation

The context of situation in which a text is embedded consists of three parties: field, mode, tenor Field is the total event, in which the text is functioning, together with the purposive activity of the speaker or writer; field thus includes the subject-matter as one element in it Mode is the function of the text in the event, including the channel taken by the language—spoken or written, extempore or prepared, and its rhetorical mode Tenor refers to the type of role interaction, the set of relevant social relations, permanent or temporary, among the participants involved

Linguistic context concerns relations within language, patterns of meaning realized by grammar and vocabulary The scholars explained that when a lexical item

occurs, it carries with its own textual history, a particular collocational environment that has been built up during text creation This will serve as the context within which the item will be embodied on this particular occasion (Halliday & Hasan, 1976, p 289)

In Halliday and Hasan (1976), the notion of cohesion was restricted only to linguistic factors that are characteristic of texts in English while admitting that the context of situation lied outside of their theory of cohesion It entails that as limited to cohesion, lexical relations described by Halliday and Hasan follow to be limited to linguistic context as well, which takes effect between cohesive pair items, rather than their surroundings This manifests right in the definition of cohesion (Section 1.1) occurring when interpretation of an element depends on another in text The scholars explained that when a lexical item occurs, it carries with its own textual history, a particular collocational environment that has been built up in the course of text creation This will serve as the context within which the item will be incarnated on this particular occasion (1976, p 289)

Lexical relations from Halliday and Hasan‘s model have been widely investigated by numerous empirical studies in EFL writing (Alarcon, 2013; Connor, 1984; Khalil, 1990; Crossley, et al., 20016; Zhang, 2000; Tran, 2011) Connor (1984) adopted the framework of cohesion to investigate the correlation between cohesion and coherence in ESL learners‘ texts These studies mainly investigated lexical relations in isolation without explicit treatment of contextual variables Thus, it is hard

to determine how these factors bear upon the analysis of lexical relations

Unlike Halliday and Hasan, Tanskanen (2006) adopted the three-partied division

of context Linguistic context refers to the language material surrounding the object being investigated: what is being produced is constrained by what has gone earlier while constraining what is to follow Cognitive context covers the cognitive factors of communication: not only mental representations and assumptions, but also the cognitive effort required from the communicators Social context is the broadest: it

includes the communication channel, the situation, the communicators, and their interactional role

Tanskanen (2006) highlighted the notion of discourse-specific relations The author provided an example to illustrate how lexical relations are specific to

discourse-the pair pausing and breather as in the following extract:

We are pausing on the road for no other reason than that we have been bounding ahead so rapidly and could all do with a breather

(Tanskanen, 2006, p 56)

The unsystematic equivalence pausing and a breather demonstrates why

classifications relying on systematic lexical relations would run into difficulty, as the items belong to different word classes, whereby indicate unsuitability of ready-made classification In general, Tanskanen focused on categories of reiteration and collocation without discerning the functional role of each type of lexical relations, for example, how synonymy contributes to communicative value

Meanwhile, Martin (1992) account of lexical relations is specific to the contextual variable field Hyponymy and hyperonymy are called by different names according to field Their typical realizations include: class, kind, type, form, breed, make, sort, style, species, order, family, variety, genre, grade, brand, caste, category These will be used as tests for hyponymy Similarly, meronymy varies from field to field, commonly realized by part, content, ingredient, fitting, member, constituent, stratum, rank, plane, element, factor, component, faction, excerpt, extract, selection, piece, segment, section, portion, measure (Martin, 1992)

The current study conforms to the contextual variables proposed by Halliday and Hasan (1976), and Martin (1992) since they are specifically relevant to lexical relations Field is determined by the topic required by writing tasks Mode pertains to written expository genre and rhetorical mode Tenor bears on learners writing to be evaluated by teacher according to prescribed criteria