A Topical Structure Analysis To Create Coherence In English Essays Of Fourth Year Students At English Department At Vietnam National University, University Of Language And International Studies.pdf

VIETNAM NATIONAL UNIVERSITY, HANOI
UNIVERSITY OF LANGUAGES AND INTERNATIONAL STUDIES
FACULTY OF POST-GRADUATE STUDIES

NGUYỄN THỊ HỒNG

A TOPICAL STRUCTURE ANALYSIS TO CREATE
COHERENCE IN ENGLISH ESSAYS OF FOURTH YEAR
STUDENTS AT ENGLISH DEPARTMENT AT VIETNAM
NATIONAL UNIVERSITY, UNIVERSITY OF LANGUAGES AND
INTERNATIONAL STUDIES

(Phân tích cấu trúc chủ đề để tạo liên kết mạch lạc trong các bài
luận tiếng Anh của sinh viên năm 4 khoa Anh, trường đại học
Ngoại Ngữ, đại học Quốc Gia Hà Nội)

M.A. MINOR PROGRAMME THESIS

Field: English Linguistics
Code: 60220201

HANOI- 2016


VIETNAM NATIONAL UNIVERSITY, HANOI
UNIVERSITY OF LANGUAGES AND INTERNATIONAL STUDIES
FACULTY OF POST-GRADUATE STUDIES

NGUYỄN THỊ HỒNG

A TOPICAL STRUCTURE ANALYSIS TO CREATE

COHERENCE IN ENGLISH ESSAYS OF FOURTH YEAR
STUDENTS AT ENGLISH DEPARTMENT AT VIETNAM
NATIONAL UNIVERSITY, UNIVERSITY OF LANGUAGES AND
INTERNATIONAL STUDIES

(Phân tích cấu trúc chủ đề để tạo liên kết mạch lạc trong các bài
luận tiếng Anh của sinh viên năm 4 khoa Anh, trường đại học
Ngoại Ngữ, đại học Quốc Gia Hà Nội)

M.A. MINOR PROGRAMME THESIS

Field: English Linguistics
Code: 60220201
Supervisor : Professor Hoàng Văn Vân

HANOI- 2016


DECLARATION
I, the undersigned, hereby certify my authority of the study project report
entitled

A TOPICAL STRUCTURE ANALYSIS TO CREATE COHERENCE
IN ENGLISH ESSAYS OF FOURTH YEAR STUDENTS AT ENGLISH
DEPARTMENT

AT

VIETNAM

NATIONAL

UNIVERSITY,

UNIVERSITY OF LANGUAGES AND INTERNATIONAL STUDIES
submitted in partial fulfillment of the requirements for the degree of Master in
English Linguistics. Except where the reference is indicated, no other person‟s work
has been used without due acknowledgement in the text of the thesis.
Moreover, I hereby declare that all the essays which have been used as the
primary source for the research has been allowed to be investigated by the students
who wrote the essays and the teacher who rated the essays and all these essays
haven‟t been used in any preceding research.

Hanoi, 2016

Nguyễn Thị Hồng

i


ACKNOWLEDGEMENTS
This thesis could not have been completed without the help and support from a
number of people.
First and foremost, I would like to express my sincere gratitude to Professor
Hoàng Văn Vân, my supervisor, who has patiently and constantly supported me
through the stages of the study, and whose stimulating ideas, expertise, and
suggestions have inspired me greatly through my growth as an academic researcher.
Secondly, I would like to give my deep thank to Mr Lê Thành A (the
pseudonym name of the lecturer), the lecturer of Literature in English in Vietnam
National University, University of Language and International Studies, and he is the

rater of the essays which I used as primary source of data for my research.
Last but not least, I am greatly indebted to my family, my friends for the
sacrifice they have devoted to the fulfillment of this academic work.

ii


ABSTRACT
The present paper sets out to investigate English essays of Vietnamese
learners using the Topical structure analysis (TSA) developed by Lautamatii (1978).
In the population of forty essays written by forth year students at ULIS, VNU,
twenty-six essays including thirteen high-rated and thirteen low-rated ones were
taken out for two objectives. The first objective is to compare and contrast the
physical features of two groups and the topical structures. The second objective is to
establish a possible relationship between the essay holistic evaluation and the type
of topical progressions used. Data analysis has revealed the moderately physical
differences between two groups in terms of t-units, topical depths, but no significant
difference could be found in the numbers of paragraphs and topical structures. In
addition, findings also indicated that there was significant difference in the
progression proportions used in two groups, especially in parallel progression in
body paragraph and the subtypes of sequential progression in introductions and
conclusions.

iii


TABLE OF CONTENTS

DECLARATION ........................................................................................................ i
ACKNOWLEDGEMENTS ....................................................................................... ii

ABSTRACT .............................................................................................................. iii
TABLE OF CONTENTS .......................................................................................... iv
LIST OF TABLES AND FIGURES ....................................................................... viii
INTRODUCTION .......................................................................................................1
1.

Rationale for the research............................................................................... 1

2.

Aims and objectives of the research ............................................................. 2

3.

Research questions ......................................................................................... 2

4.

Scope of the research ...................................................................................... 3

5.

Methods of the study ...................................................................................... 3

6.

Significance of the research ........................................................................... 4

7.

Structural organization of the thesis .............................................................. 4

DEVELOPMENT .......................................................................................................5
CHAPTER 1 - LITERATURE REVIEW................................................................5
1.1

Theoretical background of coherence in written texts ...............................5

1.1.1 Overview of coherence ........................................................................................ 5
1.1.2

Approaches to coherence ............................................................................... 6

1.1.2.1 Cohesion ...................................................................................................... 6
1.1.2.2 Genre ............................................................................................................ 7
1.1.2.3 Semantic relation......................................................................................... 7
1.1.3

Methods for examining coherence in a text ................................................. 8

1.1.3.1 Latent semantic analysis............................................................................. 8
1.1.3.2 Topical structure analysis (TSA) ............................................................... 9
1.2

Theoretical background of TSA .................................................................9

1.2.1

Topical Structures ......................................................................................... 10

1.2.2

Topical progressions..................................................................................... 11

iv


1.3

Review of the related studies ...................................................................13

1.3.1

The findings of prominent studies............................................................... 13

1.3.2

The findings of other research ..................................................................... 14

1.4

Summary ..................................................................................................15

CHAPTER 2 – RESEARCH METHODOLOGY .................................................16
2.1

Research methodology .............................................................................16

2.1.1 Restatement of research questions .................................................................... 16
2.1.2

Research approach ........................................................................................ 16

2.1.3

Principles/Criteria for intended data collection and data analysis ........... 17

2.2

Research methods .....................................................................................17

2.2.1

Topical Structure Analysis method ............................................................. 17

2.2.2

Data collection techniques ........................................................................... 18

2.2.3

Data analysis procedure ............................................................................... 19

2.3

Summary ..................................................................................................23

CHAPTER 3 - FINDINGS AND DISCUSSIONS................................................24
3.1

The physical features of essays ................................................................24

3.1.1

Findings on the essay physical features ...................................................... 24

3.1.1.1 Numbers of essay T-units, paragraphs and sub-topics ............................ 24
3.1.1.2 Findings on the topical structures of clauses .......................................... 26
3.1.2

Discussion on the physical features of essays (Research question 1)...... 27

3.1.2.1 Numbers of t-units, paragraphs, and sub-topics .......................................... 27
3.1.2.2 The topical structures of t-units in essays ..................................................... 28
3.2

Proportions of topical progressions in essays ..........................................28

3.2.1

Findings on the difference between low-rated and high-rated essays ..... 29

3.2.1.1 Topical progressions used in all essay paragraphs ................................ 29
3.2.1.2 Topical progressions in introductions and conclusions......................... 31
3.2.2

Discussion on the uses of topical progressions .......................................... 33

3.2.2.1 Topical progressions in all essay paragraphs and body paragraphs..... 33
3.2.2.2 Topical progressions in introductions and conclusions......................... 36

v


CONCLUSION .........................................................................................................41
REFERENCE ............................................................................................................44
APPENDICES……………………………………………………………….............I

vi


ABBREVIATIONS

LSA:

latent semantic analysis

TSA:

topical structure analysis

Topical S:

topical subject

Mood S:

mood subject

ISE:

initial sentence element

SP(s):

sequential progression(s)

SP(d):

directly related sequential progression

SP(i):

indirectly related sequential progression

SP(u):

unrelated sequential progression

PP(s):

parallel progression(s)

EPP(s):

extended parallel progression(s)

vii

LIST OF TABLES AND FIGURES

Table 1:

Essays‟ score description

18

Table 2:

The mean, mode, median of the numbers of T-units,
paragraphs, and topical depths of low-rated and highrated essays

24

Table 3:

The variation of the frequency of T-units in
introductory, body and concluding paragraphs of essays

25

Table 4:

The uses of three subtypes of SP in introductions

32

Table 5:

The subtypes of sequential in the conclusions

33

Figure 1:

The topical structures of clauses in low-rated and highrated essays

26

Figure 2:

The difference on the portions of each type of topical
progression in low-rated and high-rated essays

29

Figure 3:

The proportions of different types of progression in the
body paragraphs of two groups of essays

30

Figure 4:

The proportions of different types of progression in the
introductory paragraphs

31

Figure 5:

The proportions of different progression types in the
concluding paragraphs

32

viii


INTRODUCTION
1.

Rationale for the research

Coherence seems to be a key indicator of the text quality as it deals with the
global and local meanings and the deep organizational logic of the text. The more
coherent the texts are, the higher the text quality achieves. In general, evaluating the
text coherence is supposed as problematic because coherence has been considered
to be a subjective and hazy concept which is hard to learn and teach (Lee, 2002). As
Halliday and Hassan (1976) indicated, the distinction between a text and a
collection of unrelated sentence is a matter of degree; therefore, the holistic
evaluation is quite subjective in measuring text quality because it lacks reliability
and validity for not indicating particular strengths and weaknesses of learner texts.
However, subjective reader‟s evaluation on text is still a dominant method in
evaluating the text in teaching contexts.
Nowadays, a number of investigations have been carried out to examine how
coherence is achieved in learner texts and to help the teachers make sound
judgments on student compositions. One major method is to analyze the topic

development of the high-rated and low-rated essays by applying the Topical
Structure Analysis (TSA) developed by Lautamatii (1978). This method helps to
examine the links between the reader‟s evaluation and the topical developments of
the text to explicitly state the text features for classifying the text quality. Overall,
different patterns (topical progressions) of topical structure appear to provide good
predictors of student writing quality.
In general, this research investigated the relationship between the holistic
evaluation and the TSA of the student essays to find out the differences in topical
structures and the topical progressions used between low-rated and high-rated
essays.

1


2.

Aims and objectives of the research

The research population is the collection of forty essays written by forth-year
students at ULIS, VNU. The research aims at investigating the relationship between
holistic evaluation and the TSA of the essays by examining thirteen high-rated and
thirteen low-rated essays for two main objectives. The first objective focuses on the
physical features of the essays including the numbers of t-units (the shortest unit
consisting of one independent clause together with whatever dependent clauses are
attached to it), paragraphs, topical depths as well as the topical structures in the
essays. The research examines the differences between low-rated and high-rated
essays in terms of physical organization and the various combinations of the mood
subjects (mood S), the topical subjects (topical S), and the initial sentence element
(ISE) of the clauses in the essays in order to identify the contribution of these
features to essay quality. The second objective examines the relationship between

the text quality and the topical progressions used in student essays. It focuses on
examining the proportions of topical progression types (sequential, parallel, and
extended parallel) in two group essays. Particularly, it separately investigates the
proportions of progressions used in the introductions, the body paragraphs, and the
conclusions of essays to have a view on the contribution of these progressions in the
essays. To sum up, the research aims to reveal the physical features, topical
structures and progressions that contribute to the quality of the essays.
3.

Research questions

In order to achieve two main objectives, the research focuses on answering
two main following research questions:
Research question 1: How do the physical features in terms of lengths and
the topical structures reveal the differences between low-rated and high-rated essays
of Vietnamese students?
Research question 2: What is the relationship between holistic evaluation
and the topical progressions used in essays of Vietnamese students?
2


4.

Scope of the research

The research focuses on investigating the English essays written by
Vietnamese students. Particularly the research explores the essays of the forth year
students at ULIS, VNU written in their American Literature course in the year of
2013. These essays are chosen as the primary data for the research because of three
reasons. Firstly, these essays share many similarities including the same topic which

is about Buck‟s conflict in the story “The Call of the Wild”. Secondly, these essays
are commented and scored by the teacher, and sent back to the students through
email, it brings two advantages for the researcher. One is the convenience for
collecting all the data without difficulty and the other is that the essays had been
scored by the teacher; therefore they were an appropriate set of data for searching
answers for the research questions.
5.

Methods of the study

The research applies the TSA framework as developed by Lautamatii (1978)
to identifying the relationship between clauses in the paragraphs. The choice of
topical structure as the focus of this study is motivated by the following
considerations. First, topical structure analysis probes an important aspect of texts,
namely the patterns of maintenance and shifts of topics. Such patterns contribute
considerably to the text coherence, to the identification of what a particular stretch
of discourse is about, and, consequently, to the comprehensibility of texts.
Therefore, this is an objective method for analyzing the student writing. Second,
topical structure analysis allows for the quantification of data which makes the
study more reliable, because the frequency and percentage of t-units, paragraphs,
topical depths, topical structures and topical progression types in student essays are
calculated based on the carefully analyzed of essays. In summary, using TSA as a
tool for analyzing student essays is the most suitable choice for this research.

3


6.

Significance of the research

In general, the research is highly significant for the teaching and learning of
English writing in Vietnam. The identification of physical features, topical
structures and the dominant topical progression used in essays of Vietnamese
students supplies understanding on the essay writing features of Vietnamese
students. In addition, comparing the topical progressions of highly rated essays and
low rated essays has pedagogical implications for teaching students how to write a
good essay. Teachers can use topical structure analysis to help orient ESL students'
writing to the underlying structures that build coherence.
7.

Structural organization of the thesis

The research consists of three main parts, part one: Introduction, part two:
Development, and part three: Conclusion. Part one briefly describes the rationale for
conducting the research, two objectives of the research and the Topical structure
analysis (TSA) method used. Part two includes three chapters: Literature review,
Research methodology, Findings and Discussions. Chapter one focuses on the
theoretical background as well as the review of previous research with the similar
interest. In the theoretical background section, the overview, approaches as well as
the measurements of coherence are concerned. In addition, it presents the
background information of TSA. In the review of related literature, there are the
reviews of research on the proportions of topical progressions in student essays in
different contexts and their results, and of the studies which examined the difference
between low-rated and high-rated essays in terms of the use of topical progressions.
In chapter 2, research method of the study is in some detail described. It introduces
the population and sample of the research, the data analysis procedure. The third
chapter, Findings and Discussions, focuses on answering two main research
questions using the results from the data analysis. And part three summarizes what
has been found in the research, provides some implications as well as suggestions

for further research.

4


DEVELOPMENT
CHAPTER 1 - LITERATURE REVIEW
1.1

Theoretical background of coherence in written texts

1.1.1 Overview of coherence
“There is little consensus on the matter of an overall definition of coherence”
as Grabe and Kaplan (1996: 67) indicated. In general, the understanding of
coherence develops through the time with the focus from text-based features such as
cohesive devices (Halliday and Hassan, 1976) and semantic property of discourse
(Van Dijk, 1977) to the reader-based features (Johnson, 1986).
In the traditional approach, linguists including Halliday and Hassan (1976),
Van Dijk (1977) and Enkvist (1978) focus on the text-based features to explore text
coherence. On one hand, Halliday and Hassan (1976) focus on linguistic features of
the texts. According to them, coherence or texture is the feature of a text, which
identifies the text as a total unit. A text is considered coherent when the various
parts of a paragraph are connected together by cohesive ties. On the other hand,
focusing on the semantic meaning of the text, Van Dijk (1977) and Enkvist (1978)
identified coherence in a text by looking at the semantic relationship between
sentences in the text. Van Dijk (1977) indicated that coherence may be considered
as a “semantic property of discourses, based on the interpretation each individual
sentence relative to the interpretation of other sentences”. In addition, Enkvist
(1990) also defines coherence as “the quality that makes a text conform to a
consistent world picture and is therefore summarizable and interpretable”, and

coherence is primarily related to the nature and property of the text.
The understanding of coherence widens from text-based into reader-based
features by Johnson (1986: 247) who defines coherence in written text is “a
complex concept, involving a multitude of reader- and text-based features”.
According to Johns (1986: 250), coherence is “a feature internal to text, either in

5


terms of the linking of sentences (cohesion) or as the relationships among
propositions in the text (sticking to the point)”. Text-based coherence refers to an
inherent feature of the text, which has been discussed in the preceding paragraph.
Reader-based coherence, in contrast, requires successful interaction between the
reader and the text. Following Johnson‟s point of view, many linguists also defines
coherence in terms of reader-based features.
In summary, coherence includes two main sides: text-based and reader based
features. These sides co-exist and create the writing coherence.
1.1.2

Approaches to coherence

Coherence is an abstract concept. It is, therefore, defined differently by
different linguists. In this section, three aspects of coherence including cohesion,
genres, and semantic relation are focused to enlighten the scopes in which
coherence has been studied.
1.1.2.1 Cohesion
According to Halliday and Hasan (1976), cohesion is the major characteristic
of coherence. Cohesion occurs where the interpretation of some elements in the
discourse is dependent on that of another (Halliday&Hasan, 1976). Cohesion
depicts how meaning-based relationship is set up by lexical and syntactic features.

These explicit lexical and syntactic features are known as cohesive devices,
signaling the relationship in sentences and paragraphs. Halliday and Hasan (1976)
introduced five different types of cohesive devices in order to provide a guideline
for studying and judging the cohesion and coherence of writing: reference,
substitution, ellipsis, conjunction and lexical cohesion. They contended that through
analyzing the use of cohesive devices, one could evaluate or assess writing quality
from the perspective of coherence.
However, the cohesive theory proposed by Halliday and Hasan (1976) was
challenged by Carrell (1982) and Johns (1986) who argued for the importance of

6


readers‟ background knowledge. In other words, both the structure and content of
the text and the readers‟ operation on the text should be taken into consideration.
This leads to the second approach to coherence which is genre approach.
1.1.2.2 Genre
According to Hinkel (2004), coherence is defined as “the organization of
discourse with all elements present and fitting together logically”. This denotes that
a coherent essay consists of an introduction, a thesis statement, rhetorical support,
and a conclusion. In addition, Kaplan (1987); Ostler (1987); and Connor (1996)
have demonstrated that language and writing are cultural phenomena and
consequently each language tends to have its own rhetorical conventions which are
unique to it. In particular, the situation is the same when Vietnamese students write
their essays in English. On one hand, regarding the structural patterns that would
characterize Vietnamese discourse, Vietnamese writings enjoy their own sense of
“literatedness” which Kaplan (1966) has considered as an approach by indirection.
On the other hand, Farrel (1997a, 1997b) argues that school literacy in most English
countries is a highly specialized discourse that “giving directly to the point” or
being linear in its development. Therefore, the relevant patterns in Vietnamese

writing may be seen as digressive by some English readers. When conducting an
English essay, Vietnamese students need to pay attention to the rhetorical patterns
of English language. However, as far as an essay organization is concerned, both
English and Vietnamese essays share these three components: introduction, body
and conclusion. What make the difference are the things included in each part. The
current study approach to analyze the essay components separately to have deep
view on the ways ideas organized in each part.
1.1.2.3 Semantic relation
In identifying the coherence in writing, most researchers focus on the semantic
relations in the text. For example, Beaugrande and Dressler (1981:85) claimed that
“continuity of sense is the foundation of coherence, being the mutual access and
7


relevance within the configuration of concepts and relations”. Without the
continuity, any piece of writing is just plain writing, without making much sense to
the reader about the points it makes. In addition, Celce-Murcia and Olshtain (2000)
stated that the overall coherence of a longer text depends on the coherence within
each paragraph or section of the text. This textual relationship is partially a result of
coherent organization of the propositions and ideas presented in writing.
In particular, the study of „topic‟ in discourse becomes a promising branch of
research in examining the semantic relation of essays. Prague School linguists such
as Danes (1974), Firbas, (1974), and Mathesius (1975) have contributed to the
analysis of extended texts, emphasizing the role of sentences in the context of whole
texts, rather than any individual units of discourse. Following the Prague School,
Lautamatii (1987) developed TSA to examine the topic development in the
discourse. Obviously, there are many methods for identifying the semantic relation
between sentences in a text. However, two most promising methods are Latent
Semantic Analysis (LSA) and TSA. In the following part, brief reviews on these
two methods are carried out, and the reasons for choosing TSA as the primary

method in this research are stated.
1.1.3

Methods for examining coherence in a text

Researchers have proposed various techniques by which teachers can measure
coherence. Most of these techniques are based either on statistical analysis, such as
the LSA proposed by Foltz, Kintsch, and Landauer, (1998), and Lautamatti‟sTSA.
1.1.3.1 Latent semantic analysis
LSA is a method for measuring textual coherence. LSA is a fully automatic
statistical technique for extracting and inferring relations of expected contextual
usage of words in passages of discourse. Word, sentence and passage meeting
representation derived by LSA have been found capable of simulating a variety of
human cognitive phenomena, ranging from developmental acquisition of

8


recognition vocabulary to word-categorization, sentence-word semantic priming,
and judgments of essay quality and discourse comprehension (Landauer, 1998). The
method is proved to be automatic and fast, permitting quick measurements of the
semantic similarity between pieces of textual information. However, this method is
quite complicated and it requires much information about computers, the
application of this method for teaching practice is not favored. This technique is
unlikely to be employed by teachers or students because it is time-consuming and
require a considerable amount of training.
1.1.3.2 Topical structure analysis (TSA)
Besides LSA, drawing on Prague School research, Lautamatti (1987)
developed what she called TSA as a tool for analyzing the coherence of writing.
The use of TSA as a device in investigating the progressions in writing has been

validated by several researchers as a clear demonstration of style and strategy in
linking ideas within paragraphs and reflection of thought progression. In general,
the study of TSA has gained interest among researchers in their desire to determine
thematic development in paragraphs and styles of writing of individuals. The detail
information about TSA will be stated in the next part, which covers all the issues
related. Because it is the method which is applied to identify the relationship
between holistic evaluation on the quality of the essays and the topical progressions
used.
1.2

Theoretical background of TSA

TSA refers to the analyzing of coherence derived by examining the internal
topical structure of the paragraph as reflected by the repetition of key words and
phrases. In TSA, two important terms are topical structures and topical
progressions.

9


1.2.1

Topical structures

Topical structures are the combinations of the mood subject (mood S), topical
subject (topical S), and initial sentence element (ISE) of the t-units in the discourse.
Firstly, mood S is the structural subject of the clause. Its position is before the main
verb or auxiliary verb in the clause. And topical S is the lexical mood S which is
directly related to the discourse topic. In structural dummies such as there-sentence
or empty-it, the topical subjects are the noun phrases at the predicate, and these

structural dummies are non-topical subjects. In addition, non-topical subjects are
also the lexical mood S which are not directly related to the discourse topic. Nontopical linguistic materials include discourse connectives, illocution markers,
modality markers, attitude markers, metatextual and commentary markers.
Discourse connectives are linguistic items indicating order and logical connectors
like consequently, however…Illocution markers are used to make explicit the
illocutionary force of the statement concerned such as for example, to illustrate the
point. Modality markers illustrate the truth value of the information discussed
including it seems probably, obviously, somebody suggests, I doubt whether.
Attitude markers such as I would like to, it seems futile to… are used to make
explicit the writer own attitude. Metatextual and commentary markers such as next,
we shall discuss, in later chapter, we will attempt to, are used to comment on the
discourse itself, language of the text or the organization of the text. Finally, ISE
refers to the initially placed discourse material in sentence, whatever its form or
type. It may be the topical S, mood S or non-topical subject.
In general, there are five possible combinations of the mood subject, topical
subject, and the ISE in a clause, and they are abbreviated as T1, T2, T3, T4, and T5.
Type 1 (T1): ISE = mood S = topical S

Type 4 (T4): ISE = topical S ≠ mood S

Type 2 (T2): ISE ≠ mood S = topical S

Type 5 (T5): ISE ≠ mood S ≠ topical S

Type 3 (T3): ISE = mood S ≠ topical S

10


The clause belongs to T1 type if the ISE, mood S, and topical S coincide. One

example is the t-unit “The call of the wild (T1) was the best seller book which
brought fame to Jack London in 1903”. In this t-unit, the noun phrase “The call of
the wild” is the ISE, the mood S as well as the topical S of the clause, so the noun
phrase “The call of the wild” is bolded and marked (T1) after it. The “T2” type is
put after the topical S if the ISE is separated from mood and topical S, which
coincide. The t-unit “In order to live in a wild life, Buck (T2) has an external
conflict with Spitz- another sled-dog during their trip to go to Northland” is an
example. The ISE of the t-unit is the adverbial phrase “In order to live in a wild
life”, and the noun phrase “Buck” is both mood S and topical S of the clause. The
topical structure of the t-unit is marked T3 if ISE and mood S coincide while topical
S separate. There are two subtypes in type 3; in one variant, the mood S of the
initial main clause is a dummy, with the noun phrase in the predicate representing
the topical S, as in the t-unit “it is the moment (T3) that Buck aware of the law of
wild life”. The other variant, the initial main clause is non-topical subject, while the
topical S appears in a following sub-clause. For example, it is the t-unit “It seems
that Thornton’s Death (T3) releases Buck from the fragment bond with
civilization” in one essay in the sample. The topical S belongs to T4 if ISE and
topical S coincide, while mood S is separate. T4 often occurs in the t-unit which
consist of the topical S in an initial sub-clause and a cohesive anaphoric pronoun as
the mood S in the main clause, the clause “When Buck becomes a sled-dog with a
struggle for survival, he (T4) must be used to the law of club and fang day by day,
lesson after lesson” is an example. The last type, T5 occurs in the clause in which
the ISE, mood S, and topical S are all separate. For example, the clause “In
addition, it is also the moment (T5) that Buck recognizes the wildness, cold, and
ruthless in Spitz’s face”, the connective “in addition” is the ISE, empty-it is the
mood S, and “the moment” is the topical S. In summary, this part has made a brief
explanation on the topical structures and its types.

11

1.2.2

Topical progressions

In general, discourse topic is the main idea discussed in the whole essay or
text. Subtopics are the subordinate ideas to support for the discourse topic, and they
are directly related to the discourse topic. The hierarchically arrangement of
subtopics expresses the development of ideas to contribute to the discourse topic.
According to Lautamatii (1978), the way the written sentences in discourse relate to
the discourse topic and its subtopics is called topical development of discourse.
Topical progressions which show the topical development is the sequence of
developing one sub-topic, adding new information about it in the predicate of that
subtopic, and then proceeding to develop another sub-topic in order to contribute to
the development of the discourse topic. In general, Lautimatti (Ibid.) proposed three
types of topical progressions including parallel (PP), sequential (SP), and extended
parallel (EPP). PP consists of two consecutive clauses with the same topical subject.
The use of PP is perceived by the readers as repetition of the same topics in the
essay and reinforcement of the ideas, but, if used over-extensively and at the
expense of EPP, may produce the impression of redundancy. The second type is SP,
in which the topic of the next sentence is the comment of the previous sentence, or a
new different topic. And, the use of SP is perceived by the readers as development
of individual topic by adding details to an idea. If used sensibly and in good balance
with other types of progression, it may add to an overall good quality, whereas too
much development for a sentence topic may distract the readers from the main
ideas. Thirdly, the topic in EPP is the same as the topic of the previous sentence but
is interrupted by some SP. The use of EPP is perceived by the readers as good focus
and explicitness of expression. Absence or sparse use of EPP may produce an
overall impression of an unfocused text which lack global meaning.
In this research, three types of progressions in the paragraphs of the essays

were identified using the TSA method. In addition, based on research of Schneider

12


and Connor (1991), three subtypes of SPs also examined for further research on the
differences between low-rated and high-rated essays.
1.3

Review of the related studies

In general, TSA is considered as an objective method to check coherence and
proportions of topical progression in students‟ texts (Gao,2012; Almaden, 2006;
Barabas&Jumao-a, 2009; Carreon, 2006; Phuwichit (2004), Simpson, 2000;
Lautamatti, 1987; Witte &Faigley, 1981). In this part, the research on how TSA
identified the proportions of topical progressions in student texts as well as the use
of TSA as a tool for classification between low-rated and high-rated essays would
be reviewed.
1.3.1

The findings of prominent studies

This section briefly presents the findings of previous studies that investigated
the relationship between the progressions of topical structure and the assessors'
judgments of essay‟s quality. The two most prominent studies are the studies of
Schneider and Connor (1991) and Witte (1983a, 1983b). In general, both of the
research indicated that there are significant differences between the low-rated and
high-rated essays in the frequency of the three topical progressions. However, the
findings of two studies on the progressions which distinguish between low-rated
and high-rated essays are contradictory. While Witte found that low-rated essays

contained more SPs than high-rated essays, Schneider and Connor (Ibid.) found that
highly rated essays contained a high proportion of coherence-building SPs and an
EPP that helps return the essay to its main theme.
According to Schneider and Connor (Ibid.), the differences in criteria for
coding progressions led to the differences in the findings in the studies of Schneider
and Connor (1991) and Witte (1983a, 1983b). And Schneider and Connor
(1991:422) believed that "Not all sequential topics contribute equally to the
coherence of a text". It is clear that “directly related” SP which the topic of a

13


sentence is related to the comment of the previous sentence and indirectly related
SP which the topic of the new sentence is partly related to the comment of the
previous sentence (indirectly related topics) build coherence in the essays. And the
subtopics which are not directly related to the discourse topic (unrelated SPs) are
not likely to contribute to create coherence in the paragraphs. Therefore, deep
investigations to the SP by examining three subtypes of SP contribute to the
coherence of the essays especially in introductory and concluding parts.
1.3.2

Findings of other research

Besides the research of Schneider and Connor (1991) and Witte (1983a,
1983b), there are other research which focused on searching the differences
between low-rated and high-rated writings such as Connor and Farmer (1990),
Makinen

(1992),

Burneikaite&Zabiliute

(2003)Liangprayoon,

Chaya&Thep-

ackraphong (2013).
As Connor and Farmer (1990) indicated, if a writer logically makes use of SP
and make good balance with the other types of topical progression, it may add to the
overall good quality. Additionally, essays are perceived as better quality and
particularly more coherent when the use of the PP is in good balance with the use of
EPP (Burneikaite&Zabiliute, 2003). And the texts are considered as poor quality
essays when PP is over-extensive and EPP is insufficient.
Moreover, by applying TSA to short compositions written in English,
Makinen (1992) concluded that good writers have the ability to develop the topics
in their compositions more evenly across several topic levels than mid-quality
writers and especially the poor writers. The reverse is true of the lowest topical
depths. Good writers tend to return to higher topic levels at the end of their
compositions more often than the writers in the other categories. In addition, the
research on the effectiveness of TSA instruction of Liangprayoon, Chaya&Thepackraphong (2013) found that the proportions of topical progressions used in the
highly-rated essays were not different from those used in the low-rated ones.

14


Further, both successful and less successful students employed the SP the most in
the essays.
1.4

Summary

This chapter focuses on theoretical background and literature review of the
research. The theoretical background discusses the term coherence and the TSA
framework to approach to analyze the texts. On one hand, text coherence is
expressed through text quality based on reader‟s evaluation. On the other hand,
TSA includes the analysis of topical structures and topical progressions. Therefore,
this research examines the relationship between reader‟s evaluation and TSA of the
essays. Through the literature review, there are many differences in the level of
topical depths, the proportions of topical progressions, and the balance between
topical progressions to the level of coherence in an essay, so this research aims to
identify the differences between low-rated and high-rated essays by using TSA.

15