a study of correlative conjunctions as cohesive devices the upper-secondary english textbooks) = nghiên cứu các liên từ tương liên như các phương tiện liên kết văn bản liên hệ với sách giáo khoa tiếng anh

VIETNAM NATIONAL -HANOI
UNIVERSITY OF LANGUAGES & INTERNATIONAL STUDIES
DEPARTMENT OF POST-GRADUATE STUDIES







BY: V HNG QUANG





A Study of Correlative Conjunctions as Cohesive Devices
(with reference to the Upper-secondary English textbooks)
Nghiên cứu các liên từ t-ơng liên nh- các
ph-ơng tiện liên kết văn bản
(liên hệ với sách giáo khoa Tiếng Anh Trung học phổ thông).


MA MINOR THESIS


FIELD: ENGLISH LINGUISTICS
CODE: 60 22 15

HANOI-20009

VIETNAM NATIONAL -HANOI
UNIVERSITY OF LANGUAGES & INTERNATIONAL STUDIES
DEPARTMENT OF POST-GRADUATE STUDIES







BY: V HNG QUANG





A Study of Correlative Conjunctions as Cohesive Devices
(with reference to the Upper-secondary English textbooks)
Nghiên cứu các liên từ t-ơng liên nh- các
ph-ơng tiện liên kết văn bản
(liên hệ với sách giáo khoa Tiếng Anh Trung học phổ thông).

MA MINOR THESIS

FIELD: ENGLISH LINGUISTICS
CODE: 60 22 15





SUPERVISOR: Assoc. Prof. TRN HU MNH








HANOI-2009


iv
TABLE OF CONTENTS

Page
Declaration
Acknowledgements
Abstract

Table of content
CHAPTER 1. INTRODUCTION
1.1. Rationale
1.2. Aims and objectives
1.3. Scopes of the study
1.4. Methodology
CHAPTER 2. THEORETICAL BACKGROUND
2.1. Discourse and Text
2.2. Textuality, cohesion and coherence
2.2.1. Textuality
2.2.2. Cohesion
2.2.2.1. Substitution and ellipsis
2.2.2.2. Conjunction
2.2.2.3. Reference
2.2.2.4. Lexical cohesion
2.2.3. Cohesion and Coherence
2.3. Segmenting Texts into Units
2.3.1. Using the sentence as the unit of segmentation
2.3.2. Using the T-unit as the unit of segmentation
2.3.3. Using the proposition as the unit of segmentation
2.3.4. Using the F-unit as the unit of segmentation
2.4. Semantic relations
CHAPTER 3. SYNTACTIC AND SEMANTIC RELATIONS OF
CORRELATIVE CONJUNCTIONS
3.1. Introduction
3.2. The syntax of correlative conjunctions
i
ii
iii
iv

1
1
1
1
2
3
3
9
9
10
10
11
12
13
14
16
16
17
18
18
20
24

24
25

v
3. 2.1. Correlatives and their conjunctions
3.2.2. Correlative with phrasal coordination
3.2.3. Correlative with sentential coordination

3.2.4. Correlative with conjunction phrases of different syntax.
3. 2.5. Correlatives are focus particles
3.2.5. Correlative conjunctions of “not only … (but)” are used to link two
sentences.
3.3. Semantic relations of Correlative conjunctions
CHAPTER 4. CONCLUSION
4.1. Summary
4.2. Some implications for teaching and learning correlative conjunctions
4.3. Some implications for materials
4.4. Some Implications for Translation
4.5. Conclusion
REFERENCE
25
25
27
28
29
31

32
37
37
37
38
41
42
43

1
CHAPTER 1. INTRODUCTION
1.1. Rationale
‖In the world of human beings, you won't find a language by itself - the Dutch language
strolling the canals, or the English language having a nice cup of tea, or the German
language racing madly along the autobahn. You only find discourse.‖ Robert de
Beaugrande (1997: 36)
It is unimaginable of a world without language. We get through our days exchanging
various oral and written language (or, talk and text). We live by language or discourse, not
in discrete audio or visual units but in connected sound waves and orthographic forms to
which we assign meaning on the basis of our past experience with them and on the basis of
the situations in which these waves and forms are used.
Discourse analysis is concerned with the contexts in and the processes through which we
use oral and written language to specific audiences, for specific purposes, in specific
settings. We might one cannot understand language fully without looking at language use.
My research focuses on correlative conjunctions in English. I attempt to make my
description both semantic and syntactic.
There are at least three reasons why I believe it is important to focus on correlative
conjunctions. Firstly, the correlative conjunctions will enrich our potential for interpreting
the linguistic phenomena in English. Secondly, although there has been some research in
exploring conjunctions in general, little attention has been given to the study of correlative
conjunctions. Thirdly, our students have some difficulties in understanding and using
correlative conjunctions.
1.2. Aims:
The study is descriptive in nature and aims at finding correlative conjunction use is to
connect discourse segmental units with reference to the Upper-secondary English textbook
1.3. Scope of the study:
The study concerns the contrastive analysis of correlative conjunctions which are taught in
the Upper-secondary English textbook such as both … and, either … or, neither… nor, not

only … but also in English. The data for the study are from novels in English, textbooks,
and other sources.

2
1.4. Methods of data collection and analysis.
This study will be text-based. It will focus on the data of written English. A large archive
of texts of different types, including written speech, news reports, literature, legal texts,
academic texts, will be collected. Then correlative conjunctions will be extracted from
these texts and a corpus will be established.
The data were collected by choosing from novels by famous English writers in the 20
th

centuries, mostly won the Nobel Prize, from textbooks, and other sources.
The data collected will be analysed to find the bounding of correlative conjunctions in
terms of syntax and semantics.

3
CHAPTER 2. THEORETICAL BACKGROUND
2.1. Discourse and Text
The text/ discourse dichotomy originated in the early 1970s with the explosion of
Textlinguistics on the European Continent (esp. in Germany) and Discourse Analysis in
Britain. A trivial definition that might be proposed for each of these two disciplines is that
Textlinguistics is concerned with the analysis of text, and Discourse Analysis with that of
discourse. However, what appear to be straightforward definitions hide an intrinsic
problem, namely, the confusion between the notion of text and the notion of discourse.
+ the notion of discourse has usually been defined in relation to that of text, whose first
definition already date back to the mid 1960s. The confusion between the two notions has
been due to the different meanings and the relationship that linguists have attached to the
two terms. Thus, some linguists have mostly used and defined only one of the terms. For
example, the term ―text‖ only is preferably used by Hartmann (1964), Schmidt (1973),

Halliday & Hasan (1976), whereas the term ―discourse‖ only is preferred by Longacre
(1983), Brown and Yule (1983). At times they may have mentioned the other terms but
leaving it undefined. Some other linguists have used both terms in either of the following
three ways:
(1) interchangeably;
(2) by considering discourse as a type of text;
(3) or in opposition
In the later case, the discourse/ text dichotomy has mostly been identified with a
spoken/ written, process/ product and/ or language use/ abstraction of such use
opposition. As consequence, firstly, discourse has been identified with spoken
language and text with written language. Secondly, text has been considered the
product of the process of writing, whereas the more dynamic notion of discourse has
been identified with the process of text production and comprehension. And thirdly,
text has been viewed as the theoretical nation underlying the structure of the verbal
communication.
The notions of text and discourse do not have stable, uniform identity, their nature
varying not only according to the scholar but especially according to the theoretical
framework from within which the scholar approaches the definition of the terms. Thus,

4
it is possible to systematise the definitions of text according to four frameworks:
linguistic, communicative-pragmatic, cognitive and semiotic and the notion of
discourse has been mostly defined from a communicative-pragmatic, tagmemic,
cognitive and generative framework.
* Starting with the notion of text, there are four major frameworks within which the
definition of the notion has been attempted.
1- Within a linguistic framework text has been viewed as a mere succession of
sentences, i.e, of signs between certain punctuation markers. Still within the same
framework, text has also been defined as a semantic composition. For this definition
the sentence continues to be the key component to such an extent that many linguists

have defined text or discourse in opposition to sentence (cf. Beaugrande, 1979; Wirrer,
1979; Albadaleijo Mayordomo, 1981)
The idea of wholeness underlying the conception of text as a semantic composition has
been understood mainly in two ways.
*Wirrer (1979) and Albadalejo Mayordomo (1981): conceive the property of
wholeness as the result of applying the coherence component to a set of sentences.
*Hartman (1964), Koch (1965) wholeness is the outcome of certain intersentential
or cohesive relationships. In line with Koch‘s condition of wholeness, Grames (1966)
stresses that a text consists of a series of intersentential relationships which the lexical
choice is just one of them. It appears to be an anticipation of Harweg‘s (1968)
conception of text and of the notions of textual cohesion and coherence as used by
Halliday &Hasan (1967)
A special notion of text, which function as a bridge between the conception of text as a
semantic composition maintained within a linguistic framework and that of a purely
communicative unit held within a communicative framework, is the one represented by
Halliday (1973) and Halliday & Hasan (1976), who view text as a functional-semantic
concept belonging to the textual function of language. Their systemic-functional (SF)
approach to the study of language ―means, first of all, investigating how language is
used; trying to find out what are the purpose that language serves for us‖ (Halliday,
1973:7). Apart from the ideational and interpersonal functions of language, Halliday
also recognises a textual function which is ―concerned with the creation of text‖
(Halliday, 1973:107).

5
Even though in a relatively vague manner, Haliday (ibid.) only defines text, which
appears to be a structural unit related to the situation. Its structural property, which is a
common principle to definitions produced within a linguistic framework, refers to a
cohesive tie between sentences and to ―its meaning as a message‖, which is
synonymous to an FSP analysis of the sentence into a theme-rheme organisation. The
introduction into the notion of text of an element of contextual or situational relation

constitutes the bridge between a linguistic and a communicative conception of the
term. As far as the notion of discourse is concerned, it appears to be an instance of
language use in a particular situation, of which the text is its structural unit.
Within the systemic-functional model it is not until Hasan (1977) that the notion of
text becomes a communicative unit defined as ―a verbal social event‖ (Hasan,
1977:233) and characterised, firstly, by its property of texture (i.e., ―linguistics
cohesion within the passage‖ (Hasan, 1977:228)), which constitutes a means of
differentiating it from a random chain of sentences; secondly, by its structure, which
serves to ―distinguish between complete and incomplete texts on the one hand, and
between different generic form on the other‖ (id.:229); and last but not least, by its
contextual relation. Following Halliday‘s social perspective on language analysis,
Hasan emphasises the role that context plays in the structure organisation (structure
formula) of each ―genre of text- i.e. type of discourse‖ (ibid.).(Note: Hasan (1977) uses
―text‖ and ―discourse‖ interchangeably)
The notion of context of situation in Hasan (1977) is explained through that of text
genre or register. Register is related to systematic variation in language, this variation
depending on the selection of different linguistic as well as contextual varianles. Field,
tenor and mode of discourse are the variables that constitute the contextual construct
(CC). The definition of text as a verbal social event is directly related to the three types
of roles which the interactants adopt in a communicative situation and which are
integrated in the variable tenor. These roles are:
(1) textual, which classifies the interactants into speaker and hearer.
(2) Social, which establishes a hierarchical or non-hierarchical relationship
between the interactants according to their social status; and
(3) Participatory, which identifies the initiator and the respondent of the
communication.

6
Hasan‘s (1977) conception of text as a social event would undergo an evolution from a
primarily linguistic-centred approach which defined text as a semantic unit occurring in a

situational context and whose sentences are tied by a relationship of cohesion, towards a
more communicative-centred position located within a social-semiotic approach to
language study. Though text continues to be essentially a semantic unit, it is no longer
viewed as a mere product but also a process.
- Hasan‘s (1977) and Halliday &Hasan‘s (1985) conception of text has directly led into the
second framework of approach to the notion, namely, the communicative-pragmatic
framework. For linguists within a communicative-pragmatic framework text is no longer a
succession of sentences but of ―propositions‖ or semantic units referring to events, actions
or states which contribute to a communicative situation or interaction. A proposition may
consist of a single word (e.g. greeting, farewell, addressing form), of an elliptical sentence
(e.g. verbless sentence) (note: the term ―elliptical‖ is used in Halliday (1985)), or it may
coincide with a sentence boundary. Indeed, as happens within the linguistic framework, the
sentence continues to be the most complex unit that structures information contained in a
communicative activity. But, in contrast to the previous framework, a sentence is not only
a component of a text but it may also be a complete text on its own.
One of the most outstanding textlinguists to urge for a pragmatic approach to the notion of
text is van Dijk (1977), who equates the text/ discourse opposition with a theoretical notion
vs. observation dichotomy. A grammar, in his view, should not only describe an expression
in terms of its internal structure and the meaning assigned to it, but also in terms of the
condition that render the expression acceptable in a particular communicative context. This
principle should apply not only to sentences but also to discourse. Similarly, Beugrande
&Dressler (1981) also consider the pragmatic condition of acceptability to be, along with
some others, a key feature of any text. Beaugrande &Dressler‘s (1981) communicative
approach to text analysis provides an innovation to Textlinguistics, namely, a
comprehensive description of the pragmatic components that transform a text into a verbal
interaction located in a specific situational context, with interlocutors observing certain
conversational principles necessary to fulfilment of the intended goal of the encounter. In
this framework, text is equated with an interactional process.
- A third framework which also uses the unit of text as object of study is the cognitive.
Cognitive text models work with the concept of text or discourse as a natural unit of

7
language which consists of a string of successive sentences- or utterances in spoken form-
with topical or logical structure. In other words, text or discourse is conceived as a
semantic unit forming a coherent and cohesive structural whole independent of the context
in which it is produced, assuming that the semantic structure, that is, ―the formal
reconstruction of what is non-technically called the “information” or “content” of a
discourse‖ (Van Dijk & Kintsch, 1979: 67), is the basis for all particular meanings. This
conception is similar to the one maintained by scholars within the linguist framework
explained earlier. What differentiates then the notion of text or discourse as used by one
and the other framework is that for linguistic models text is a final product, where as for
cognitive model text or discourse is conceived as a process of production, understanding,
organising and retrieval.
- The fourth and last framework within which text is studied is the semiotic one as
represented by Petofi (1977, 1980). Within this framework, text is a broad notion referring
to the unit of analysis of any sign system. Text as a semiotic object comprises both the
natural language text (vid. Petofi, 1977), also called discourse, and texts of another
semiotic character (e.g. animal communication, theology, film analysis, advertisements,
etc.).
2. From the perspective of the theoretical framework, it is possible to organise a
classification of the concept of discourse into four mains groups: communicative-
pragmatic, tagmemic, cognitive and generative. As far as cognitive framework is
concerned, suffice it to say that the discussion on the notion of text provided above is
equally applicable to that of discourse since the terms text and discourse are used
interchangeably within this framework.
- Scholars working within a communicative-pragmatic framework all agree that discourse
is language in use, that is, a unit of communication located within the wider context of
purposeful speech behaviour where the pragmatic component plays a central role ( Van
Dijk 1977, 1979; Edmondson, 1981; Brown &Yule, 1983; Leech, 1983; Levinson, 1983).
The text/ discourse opposition within this approach seems to correspond to the

competence/performance or use/ usage dichotomies. For example, Brown & Yule (1983)
consider discourse as ―language in use‖ which includes ―the purposes or functions which
those [linguistic] forms are designed to serve in human affairs‖ (brown &Yule, 1983: 1).
Discourse is viewed as a process connected to human behaviour in which producers and

8
receivers interact. Text for its part, which comprises both a spoken and written form, is its
representation, ―a technical terms, to refer to the verbal record of a communicative act‖
(Brown &Yule, 1983:6).
- A change in the geographical environment and in the theoretical approach serves to locate
discourse as an object of investigation in America, within the tagmemic framework
represented by Longacre (1979, 1983). Tagmemics, which results from an integration of
Linguistics and Anthropology, goes beyond the sentence and text boundary and locates
language within human activities. Verbal behaviour is considered to be just one strand of
man‘s activity. As a consequence of this anthropological approach to language, Longacre
view discourse as a sociological entity which may appear in the form of either ―a
conversation between two people, a planned interview on the radio, a news report, a
sermon, a political speech, a short story, an essay, a fairy tail, or a novel‖ (Longacre,
1979: 258). It refers both to monologue and to dialogue. Discourse, he continues, has
textual (as defined by Halliday& Hasan, 1976), is constituted by elements of a lower rank
level and occurs in a sociolinguistics setting: ―it has a speaker or a writer, and it is
directed at a hearer or an audience of some sort‖ (ibid.)
Also in America, the last framework to adopt the notion of discourse as its object of
study is the generative as represented by Kuno (1987). Within the domain of Generative
Discourse Analysis, discourse is understood in a similar fashion to the early conceptions of
text, that is, as a linguistic unit consisting of a chain of sentences. In this view, the
contextual factors of discourse are limited to the linguistic ―co-text‖, i.e., the preceding and
following sentences, an analysis of discourse consists then in the study of syntactic,
semantic and pragmatic principles controlling the usage of linguistic phenomena that
exclusively belong to the realm of the sentence. Thus, the study of discourse comprises

those aspects of linguistic research which generative grammarians have considered to be
outside the domain of generative syntactic theory, that is, the study of pragmatics and the
correlation between syntactic and semantic phenomena.

9
2.2. Textuality, cohesion and coherence
2.2.1. Textuality
In the approach to text linguistics by de Beaugrande & Dressler (1981), text is established
as a communicative occurrence, which has to meet seven standards of textuality. If any of
these standards are not satisfied, the text is considered not to have fulfilled its function and
not to be communicative.
Cohesion and coherence are text-centred notions, designating operations directed at the
text materials. Cohesion ―concerns the ways in which the components of the surface text
i.e. the actual words we hear or see are mutually connected within a sequence‖ (de
Beaugrande & Dressler 1981:3). Coherence on the other hand ―concerns the ways in which
the components of the textual world, i.e. the concepts and relations which underlie the
surface text are mutually accessible and relevant‖ (1981:3-7).
The remaining standards of textuality are user-centred, concerning the activity of textual
communication by the producers and receivers of texts:
Intentionality concerns the text producer‘s attitude that the set of occurrences should
constitute a cohesive and coherent text instrumental in fulfilling the producer‘s intentions.
Acceptability concerns the receiver‘s attitude that the set of occurrences should constitute a
cohesive and coherent text having some use or relevance for the receiver.
Informativity concerns the extent to which the occurrences of the text are expected vs.
unexpected or known vs. unknown/uncertain.
Situationality concerns the factors which make a text relevant to a situation of occurrence.
Intertextuality concerns the factors which make the utilisation of one text dependent upon
knowledge of one or more previously encountered texts.
The above seven standards of textuality are called constitutive principles.They define and
create textual communication as well as set the rules for communicating. There are also at

least three regulative principles that control textual communication: the efficiency of a text
is contingent upon its being useful to the participants with a minimum of effort; its
effectiveness depends upon whether it makes a strong impression and has a good potential
for fulfilling an aim; and its appropriateness depends upon whether its own setting is in
agreement with the seven standards of textuality (de Beaugrande & Dressler 1981:11).


10
2.2.2. Cohesion
Cohesion is the first of the seven textuality standards identified by de Beaugrande &
Dressler. It has also been a most popular target for research, probably because it is easy to
identify in written texts, which are the traditional research material of linguists. This does
not mean, however, that there would be a general consensus as to the definition of the
concept and its relation to the second of the textuality standards listed above, coherence.
Since cohesive markers are important for the understanding of oral texts as well as
written, it seems feasible to describe this textuality standard in some detail. Interpreters, as
all speakers, make extensive use of cohesive devices, for example in order to enhance
coherence, but also for reasons of economy (e.g. saving time and alleviating conceptual
work load by using anaphoric devices like generalisations and pro-forms).
Halliday and Hasan, in their ground-breaking work "Cohesion in English" (1976), describe
cohesion as a semantic concept that refers to relations of meaning that exist within a text.
According to Renkema (1993) cohesion is the connection which results when the
interpretation of a textual element is dependent upon another element in the text.
According to Schiffrin (1987:9) cohesive devices are "clues used by speakers and hearers
to find the meanings which underlie surface utterances".
Halliday and Hasan (1976) define two general categories of cohesion: grammatical
cohesion (substitution, ellipsis, conjunction, reference) and lexical cohesion.
2.2.2.1 Substitution and ellipsis
One type of grammatical cohesion is substitution, which takes two forms: a) substitution
per se, which is "the replacement of one item by another", and b) ellipsis, in which "the

item is replaced by nothing" (Halliday and Hasan 1976:88). There are three types of
substitution: nominal, verbal and clausal.
(a) substitution per se, (b) ellipsis (zero-replacement)
Substitution of noun:
a) These biscuits are stale. Get some fresh ones.
b) These biscuits are stale. Those are fresh.
Substitution of verb
In English, this is done by replacing a verbal expression with the lexical item "do":
a) A: Have you called the doctor?

11
B: I haven‘t done it yet, but I will do it.
A: Though actually, I think you should do it.
b) He participated in the debate, but you didn‘t.
Substitution of clause is accomplished by using the lexical items "so" and "not":
a) A: Are they still arguing in there?
B: No, it just seems so.
b) Who wants to go shopping? You?
(Examples are from Renkema 1993:37-38).
2.2.2.2 Conjunction
Conjunction is a relationship indicating how the subsequent sentence or clause should be
linked to the preceding or the following sentence or parts of sentence. This is usually
achieved by the use of conjunctions. Frequently occurring relationships are addition,
causality and temporality.
The relationship can be hypotactic, combining a main clause with a subordinate clause or
phrase, or paratactic, combining two main clauses.
*Junction
De Beaugrande & Dressler (1981) prefer to call the type of cohesion in question
"junction", and discuss four major types of junctive expressions:
Conjunction links things which have the same status, e.g. both true in the textual world.

Conjunction is the default junction, since, unless specified otherwise, events and situations
are combined additively in a text. Thus, there is no motive to place "and", "also", "in
addition" etc. between every clause or sentence, but only when interdependency is not
obvious and should be stressed (1981:71-72).
Disjunction links things which have alternative status, e.g. two things of which only one
can be true in the textual world. "Or" is the most common disjunction signal, sometimes
expanded to "either/or", "whether/not" etc. Within a sentence, "or" joins alternatives both
of which are current in active storage, but only one of which obtains in the textual world.
Between sentences, it tends rather to announce an afterthought, an alternative not
considered before. When processing disjunctions, text users will have to carry forward
both alternatives in active storage until a resolution is at hand, probably making
disjunctions difficult to process (1981:71-72).

12
Contrajunction links things having the same status but appearing incongruous or
incompatible in the textual world, e.g. a cause and an unanticipated effect. It is signalled by
"but" (most often), "however", "yet", nevertheless", etc. It is the function of contrajunction
to ease problematic transitions at points where seemingly improbable combinations of
events or situations arise (1981:71-73).
Subordination links things when the status of one depends on that of the other, e.g. things
true under certain conditions or for certain motives (precondition/event, cause/effect, etc.).
It is represented by a large number of conjunctive expressions: "because", "since", "as",
"thus", "while", "therefore", etc. Subordinating junctives make explicit
a) coherence relations, e.g. cause (necessary conditions), reason (rational human
reaction);
b) relations of temporal proximity ("then", "next", "before", "after", "since",
"whenever", "while", "during", etc.);
c) modality, i.e. the probability, possibility, or necessity of events and situations,
e.g. "if".
(de Beaugrande & Dressler 1981:71-74).

2.2.2.3 Reference
Reference is another well researched area within linguistics. It is defined by Halliday &
Hasan (1976:31) as a case where the information to be retrieved is the referential meaning,
the identity of the particular thing or class of things that is being referred to. The cohesion
lies "in the continuity of reference, whereby the same thing enters into the discourse a
second time."
In other words, reference deals with semantic relationship. Reference can be accomplished
by
 exophoric reference, which signals that reference must be made to the context of the
situation;
 endophoric reference: reference must be made to the text of the discourse itself; it is
either anaphoric, referring to preceding text; or cataphoric, referring to text that follows.
Halliday & Hasan (1976) describe the following types of reference:
 personal reference: nouns, pronouns, determiners that refer to the speaker, the addressee,
other persons or objects, or an object or unit of text;

13
 demonstrative reference: determiners or adverbs that refer to locative or temporal
proximity or distance, or that are neutral;
 comparative reference: adjectives or verbs expressing a general comparison based on
identity, or difference, or express a particular comparison.
2.2.2.4 Lexical cohesion
Lexical cohesion does not deal with grammatical or semantic connections but with
connections based on the words used. It is achieved by selection of vocabulary, using
semantically close items. Because lexical cohesion in itself carries no indication whether it
is functioning cohesively or not, it always requires reference to the text, to some other
lexical item to be interpreted correctly. There are two types of lexical cohesion: reiteration
and collocation.
Reiteration includes (examples below are from Renkema 1993)
repetition (often involving reference)

A conference will be held on national environmental policy. At this conference the issue of
salination will play an important role.
synonymy (often involving reference)
A conference will be held on national environmental policy. This environmental
symposium will be primarily a conference dealing with water.
hyponymy (superordinate vs. subordinate concepts)
We were in town today shopping for furniture. We saw a lovely table.
metonymy (part vs. whole)
At its six-month check-up, the brakes had to be repaired. In general, however, the car was
in good condition.
antonymy
The old movies just don‘t do it anymore. The new ones are more appealing.
Lahdenmäki (1989) calls these relations "(direct) synonym-type relations, since they all
refer to another word which has the same referent (e.g. ‘I met a man yesterday. The bastard
stole all my money‘)".
Collocation is any pair of lexical items that stand to each other in some recognisable
lexico-semantic relation, e.g. "sheep" and "wool", "congress" and "politician", and
"college" and "study".

14
Red Cross helicopters were in the air continuously. The blood bank will soon be
desperately in need of donors.
The hedgehog scurried across the road. Its speed surprised me.
(Examples above from Renkema 1993.)
Like in the case of synonymous reference, collocational relation exists without any explicit
reference to another item, but now the nature of relation is different: it is indirect, more
difficult to define and based on associations in the reader‘s mind (e.g. ‘I looked into the
room. The ceiling was very high.‘). Interpretation of such relations is completely based on
the knowledge of subject fields (Lahdenmäki 1989).
2.2.3. Cohesion vs. coherence

The term "cohesion" is often confused or conflated with "coherence". But it is necessary,
both from a theoretical and a practical point of view to retain this distinction between
connectivity on the surface and connectivity of underlying content.
The term coherence, apart from being polysemic, is also controversial. While de
Beaugrande & Dressler (1981) treat coherence as number two of the two text-centred
standards, Carstens (1997), in his thorough work on Afrikaans text-linguistics, takes up
coherence as the last standard of textuality, as coherence in his opinion entails all of the
other six standards. According to Lundquist (1989:123; cited in Carstens 1997) coherence
is not a typical linguistic problem, but a general principle for the interpretation of all
human activity, verbal or non-verbal. Neither is coherence a property which is inherent in
texts, but rather a property which is assigned to a text by its reader. To put it differently:
texts are not automatically coherent, but become coherent when the recipients of the texts
find them coherent (Carstens 1997:481-482).
Lahdenmäki (1989) underlines, that coherence is a purely semantic property of discourse,
while cohesion is mainly concerned with morpho-syntactic devices in discourse. A
coherent text is a semantically connected, integrated whole, expressing relations of
closeness, e.g., causality, time, or location between its concepts and sentences. A condition
on this continuity of sense is that the connected concepts are also related in the real world,
and that the reader identifies the relations. Each sentence must also "satisfy" the text topic
(van Dijk 1977:138) which "controls" or places limits upon things a concept can be related
to (de Beaugrande 1987). Therefore, if two concepts are logically and associatively too

15
distant in semantic space, they cannot function coherently, even if they were connected in
the surface text by overt cohesion markers, e.g. connectives. Instead, in a coherent text,
there are direct and indirect semantic referential links between lexical items in and between
sentences, which the reader must interpret (Lahdenmäki 1989:27).
In the present study we are not primarily interested in whether or not coherence is a purely
text-centred standard of textuality. But from a communicative point of view - because
interpreters are paid to communicate! - a text must be coherent enough for the interlocutor

to be able to interpret. It seems probable that this coherence can be achieved either through
cohesion, i.e. markers/clues in the speakers' text, or through the employment of the "user-
centred" textuality standards of intentionality, acceptability, informativity, situationality
and intertextuality.
To sum up this section, it is interesting to quote the following point made by van Dijk &
Kintsch (1983):
"On full analysis there are probably few surface structure items that are not produced in
order to signal a semantic, pragmatic, cognitive, social, rhetorical, or stylistic function.
Thus, at this level, little is left of the old Saussurian arbitrariness in the relations between
expressions (signifiers) and their meanings (signifieds)." "Nearly all underlying
(semantic, pragmatic, etc.) information can be mapped onto surface structures and parallel
paratextual action." (Dijk & Kintsch 1983:285)
But the relation between surface structures and their semantic, pragmatic, or interactional
functions on the one hand, and their relevance for production on the other, cannot be too
strict:
"Some languages have quite varied surface structures, and it remains to be seen whether
this will always directly presuppose different comprehension and production strategies."
"Further work regarding these relationships between the (functional) structures of
sentences in different languages and their cognitive processing is necessary — especially
taking into account the textual relevance of these functions." (Dijk & Kintsch 1983:285)

16
2.3 Segmenting Texts into Units
The first step in characterizing the discourse structure of a text in our studyl is to determine
the elementary discourse units (EDUs), which are the minimal building blocks of a
discourse tree.
There are many methods for segmenting text. Three of them are most commonly used,
which involve the use of
 a grammatical unit (the sentence or the T-unit)
 a semantic unit (the proposition), or

 a grammatical/ semantic unit (F-unit)
as the basic unit of segmentation. In the following section, each of these method wull be
introduced and evaluated, so as to find the method used in the present study.
2.3.1. Using the sentence as the unit of segmentation.
The sentence might appear to be an appropriate segmentation unit because it is easily
identified as a syntactic structure. Halliday and Hasan (1976) claim that the orthography
sentence, that is, whatever occurs between full stops, provides a good basis for defining
cohesion. For them, ―cohesive ties between sentences stand out more clearly because they
are the ONLY source of texture, whereas within the sentence there are the structural
relations as well‖ (Halliday and Hasan, 1976:9)
However, this definition of cohesion in terms of the orthographic sentence is not
satisfactory. For example, it is tended to reject the conjunction but in (a) and yet accept it
in (b) as cohesion, although the word certainly has the same function.
(21)The child liked the banana, but he disliked the apple.
(22)The child liked the banana. But he disliked the apple.
(23) The child liked the banana though he disliked the apple.
McCarthy (1991:153) points out:
―In all our discussions on speaking, the sentence was dismissed as being of dubious value
as a unit of discourse (especially in Chapter 4). The sentence is more obvious as a
grammatical unit in writing, although certainly not in all kinds of writing: signs and
notices, small ads, notes, forms, tickets, cheques, all contain frequent examples of 'non-
sentences' (lists of single words, verbless clauses, etc.). The internal construction of the
sentence has always been the province of grammar, but in Chapter 2, we argued that a

17
number of things in clause and sentence grammar have implications for the discourse as
a whole, in particular, word order, cohesion, and tense and aspect. For the purposes of our
discussion of these discoursal features, the sentence will have no special status other
than as a grammatical and orthographic unit which can be exploited where desired for
pedagogical illustration, just as the clause can.‖

Another reason is that some students are weak in punctuation, and overuse of cohesive
devices. And the length of the orthographic sentence is in principle unlimited.
2.3.2. Using T-unit as the unit of segmentation.
Hunt (1970) introduces the T-unit as a solution to the problem. He defines the T-unit as
―one main clause plus any subordinate clause or nonclausal structure that is attached to or
embedded in it‖ Hunt (1974:4), and argued that segmenting a text into T-units means
segmenting it into its shortest grammatically feasible units. For example (from Hunt
(1983:101):
I like that movie we saw about Moby Dick the while whale the captain said if you can kill
the white whale Moby Dick I will give this gold the one that can do it and it is worth
sixteen dollars they tried and tried but while they were trying they killed a whale and used
the oil for the lamps they almost caught the white whale.
According to Hunt, this long sentence consists of six T-units:
1. I like that movie we saw about Moby Dick the while whale .
2. the captain said if you can kill the white whale Moby Dick I will give this gold the
one that can do it.
3. and it is worth sixteen dollars
4. they tried and tried
5. but while they were trying they killed a whale and used the oil for the lamps
6. they almost caught the white whale.
The T-unit has been used in a number of studies for its advantages. Firstly, it can be
identified objectively, in terms of simple and complex sentences, or coordinated sentences
with subject deletion. Secondly, it is not affected by poor punctuation. Lastly, since T-unit
length incorporates the mean words per clause and the number of clauses per T-unit, it is a
useful index of syntactic complexity.

18
However, T-unit also has a number of weaknesses. It cannot fully solve the problem raised
in example (21)-(23). It puts (21) and (22) on the same level since both of them have two
main clauses while (23) is considered as one T-unit because it has only one mains clause

and one dependent clause. Besides, it is a large unit, it is not useful in analysing significant
rhetorical relations in the text, such as contrast, if these relations are attained by means of a
subordinator. For the reasons, the T-unit was not chosen as our segmentation unit.
2.3.3. Using the proposition as the unit of segmentation
A proposition is a statement expressing a judgement or opinion. Crombie (1985:13)
explains that the semantic relation involved in text analysis are, minimally, two
propositions. Each member of a semantic relation is often encoded as a separate clause,
since this is frequently the linguistic unit used to encode a single proposition. However, a
semantic relation may also be encoded as a group of clauses. For example:
John‘ playing squash and Mary‘s weeding the garden while Tim‘s chopping wood and
Sam‘s preparing the dinner but Jane isn‘t doing anything.
Alternatively, a semantic relational member may be encoded as a proposition embedded in
a single clause, for example:
Her exaggerations make him furious.
i.e. she exaggerates (reason)
He becomes furious (result)
→ Reason- Result
2.3.4. Using the F- unit as the unit of segmentation
For Lindeberg (1988), the most appropriate unit of segmentation is the functional unit
discourse, or F-unit. The F-unit involves using both grammatical and semantic notions in
text segmentation. It is defined as (the set of clause or clause equivalent serving an
identifiable rhetorical function in written discourse‖ (Lieber 1979: Abstract :i). Lieber‘s
segmentation principles (1979:93-95) are presented below:
1. Coordinate structures
Full clauses joined by coordinating conjunctions or marks of punctuation constitute
separate F-units.
e.g. a. John broiled the steak,
b. and Bob made the salad.

19

Clauses exhibiting gapping in a non-initial member constitute separate F-units.
e.g. a. Nick prepared the main course,
b. and Tony the desert.
Clauses containing conjoined verbal structures will be segmented into more than one F-
unit.
e.g. a. He finished his work
b. and left immediately.
Conjoined nonverbal elements within a clause will be segmented into separate F-units
when an overt marker indicating a change of rhetorical function is present (i.e. but, except
or an adverbial marker or prepositional phrase)
e.g. a. she is highly qualified
b. and therefore suitable for the position.
2. subordinate structures
Adverbial subordinate clauses and clause equivalents, with the exception of temporal
and locative structures.
e.g. a. They had to hire new teachers
b. (in order) to handle the expected increase in enrollment.
Non-restrictive relative clauses and sentence relatives
e.g. a. I met Tom‘s father,
b. who works at the university.
Non-restrictive appositives (i.e. reduced non-restrictive relative clauses)
e.g. a. The latest procedure can be found in Professor Wirth‘s new book,
b. Airport Management in Developing Countries.
Non-restrictive appositives of exemplification, identification and renaming
e.g. a. He‘s made all his arrangements with the new managing company,
b. namely, Walter and Samuels.
Absolute constructions related to adverbial clauses or non-restrictive relatives.
e.g. a. The day being sunny,
b. he decided to play golf.
The F-unit is a suitable unit for discourse segmentation since it is defined not by

grammar alone, but by rhetorical function as well.

20
Therefore, the F-unit is used in this study since it is an essential measure for the
analysis of cohesion and relation coherence.
2.4. Semantic relations
Crombie‘s set inter-propositional general semantic relations- her set of binary textual
relations- has been applied in this study. The framework of these relations was applied in
this study as shown below. The first unit is called X, and the second unit to which it relates
is called Y. A full explanation of these relations and their applicable variation now follows.
2.4.1. Temporal relation
These deal with temporal links between F-units.
(a) Chronological sequence
The event specified in Y follows the event specified in X without necessarily being
causally related to it.
I went to the child/ and knelt down before her.
(b) Reverse chronological sequence
In this case the event in Y precedes the event specified in X without necessarily being
causally related to it.
Before the bus could stop,/ everyone was running after it.
(c) Temporal overlap
The event specified in Y overlaps, either wholly or partly, in time with the event
specified in X.
While we were still waiting for the bus, / the sound started again.
2.4.2. Matching relation
These refer to some kind of similarity or contrast between F-unit segments.
(a) Comparison
Involves the comparison of two things, events or abstractions in terms of some particular in
respect of which they are similar.
At that time my friend stood, / just like a prisoner given a life sentence.

(b) Contrast
Involves the comparison of two things, events or abstractions in terms of some particular in
respect of which they differ.

21
A person in a city can start a business,/ unlike in rural areas where people usually
depend on their agricultural produce.
2.4.3. Cause-effect relations
These involve various kinds of causal relations between units.
(a) Condition-consequence
Here some aspects of X provide a condition for some aspect of Y
If you have a sister, / she will help you.
(b) Denied Consequence
Some sequence that would normally follow from a condition expressed in X is denied
in Y.
In a city, even if someone is alert to danger, / he cannot be completely safe.
(c) Reason- Result
X provides a reason for some effect of result in Y.
Because we were talking loudly, / the baby started crying.
(d) Means- Result
X explains how some result or effect is achieved in Y
By living in a city, / you can end up losing you culture.
(e) Grounds- Conclusion
X provides an observation in terms of which a conclusion is made in Y.
People in cities come from different areas. / This explains why some can be influenced
to forget their culture.
(f) Means- Purpose
Y provides a purpose, an intention, or an event in X.
One morning I went to the nearby shop/ to buy a newspaper.
2.4.4. Truth and Validity relations.

(a) Statement- Affirmation
Y affirms the truth of X.
She might have forgotten her passport. / Really she had forgotten it.
(b) Statement- Denial
Y denies the truth of X.
He says that money is better than honour, / but I don‘t agree.