PART A: INTRODUCTION
1
1. Rationale
There has been much written in recent years about Critical Discourse Analysis (CDA for
short). It appears to be quite difficult to define CDA in simple terms. However, CDA is my
choice for the MA thesis because first and foremost, CDA regards language as a social
practice – language is a part of society, language is a social process, and language is a
socially conditioned process. These implications have been discussed in details by Norman
Fairclough (1989).
Then, it can be inferred from above implications that doing discourse analysis involves in
not merely analyzing texts, processes of production and interpretation, but also analyzing
the relationship between texts, processes, and their social conditions – or in other terms,
the relationship between texts, interactions, and contexts. And CDA is critical in the sense
that it shows connections and causes which are hidden – such as the connection between
language, power and ideology, the problems of inequality and racism – through discourse
analysis. More clearly stated, CDA is critical when it explicitly addresses social problems
and seeks to solve social problems through the analysis and accompanying social and
political action. The intention of the analysts in this view of ‘critical’ is explicitly oriented
toward locating social problems and analyzing how discourse operates to construct and
historically constructed by such issues. They must work from the analysis of texts to the
social and political context in which the texts emerge. CDA thus can be said a very
practical form of discourse analysis. It seeks not only to describe language but also to offer
critical resources to those wishing to resist various forms of power.
For those mentioned reasons, a speech by Martin Luther King “Beyond Vietnam – A Time
to Break Silence” is to be analyzed from the CAD viewpoint together with the hope that
this study may vice versa help illustrate and clarify CDA concepts. Despite being delivered
in 1967, this speech was especially mentioned by many Americans when the U.S.
government decided to attack Iraq in 2003. It is the ideological and topical features of this
speech that “Beyond Vietnam – A Time to Break Silence” by Martin Luther King has been
chosen. During the course of analyzing the speech, the way power and ideology embedded
in texts and coded in language use will be gradually manifested. Although I am a supporter

• How are power and ideology realized in terms of transitivity
and thematic structures?
• How are power and ideology realized macro structurally?
3
Besides, this study also aims at:
- Providing an objective view as a linguist when approaching texts, especially those
addressing power and ideology.
- Trying to develop the ability in critical thinking.
- Finding the application of CAD approach in language teaching and learning.
4. Design of the study
The study consists of three parts. They are:
Part A: Introduction: This part presents the rationale, scope, aims, methodology, and
design of the study.
Part B: Development: This is the main part and it consists of three chapters.
Chapter 1: Theoretical background.
This chapter gives an overview of CDA – its history, role, concepts, and procedure.
Systemic-functional theory is also concerned in this chapter.
Chapter 2: A critical discourse analysis of Martin Luther King’s speech: ‘Beyond
Vietnam – A Time to Break Silence’.
The CDA procedure addressed by Fairclough and systemic-functional theory are applied to
analyze the speech to find out the connection between power/struggle, ideology and the
language.
Chapter 3: Implications
Some implications of applying CDA in developing critical thinking, in language teaching
and learning.
Part C: Conclusion
This part summarizes the main findings of the study, draws important conclusion and
offers suggestions for further research.
References
Appendices

macrostructure or overall idea of the speech, the text will be divided into major sections
and then four procedures namely attributive deletion, predictive deletion, simple
5
generalization, and integration suggested by Van Dijk (1977, 144-146) will be applied. In
this part, my aim is to uncover power and ideology hidden behind words.
The next step is interpreting the relationship between the processes of text production and
interpretation. There are two domains here: interpretation of situation context and
interpretation of intertextual context. In the interpretation of situation, I follow questions
given by Fairclough (2001): what’s going on, who’s involve, what relationships are at
issue, and what’s the role of language. In interpretation of intertextual context,
presuppositions are in focus.
And then, explanation concerned with the relationship between those processes and social
context – how the discourse change or sustain certain social relationship in social structure
– is needed. The speech will be looked at as a social practice in relation with other social
practices.
Finally, implications will deduced from my understanding in CDA, in language teaching
and learning methodologies as well as my own experience.
6. Significance
Theoretically, this study provides a support to CDA theories. From an objective view as a
linguist when approaching texts, CDA analysts can find out ideology and power hidden
behind words. Practically, this study is submitted in partial fulfillment of requirements of
my degree of Master in Linguistics. Moreover, it may provide me another approach to
language teaching and learning: looking at language teaching and learning from CDA
viewpoint.
6
PART B: DEVELOPMENT
7
Chapter 1:
THEORETICAL BACKGROUND
1.1. An overview of CDA

merely a reflection or expression of social processes and practices, it is a part of those
processes and practices. For example, disputes about the meaning of political expressions
are an aspect of politics.
Secondly, language is a social process. Language comes into life in form of texts (the term
Michael Halliday uses for both written and spoken texts). In CDA, text is not discourse.
Text is the product of text production and the resource for the process of interpretation.
During the processes of text production and text interpretation, people have to depend on
what they have in minds – including their knowledge of language, natural and social
worlds, values, beliefs, assumptions and so on. In other terms, text is the traces of the
productive process and cues in the interpretative process. (Fairclough, 2001)
Thirdly, language is a socially conditioned process as the processes of production and
interpretation are socially determined. Text producers and interpreters not only draw upon
what are there in their heads, they are also socially generated and socially transmitted.
People internalize what is socially produced and made available to them and use this to
engage in their social practice, including discourse. To make it clearer, Fairclough (2001)
calls what discourse participants have in their minds during the course of interpreting and
producing texts as Member’s Resources (MR) - resources for productive and interpretative
processes. MR have both cognitive and social features as they come from people’s mind
while they are socially originated.
Thus, when CDA sees language as discourse and as a social practice, apart from analyzing
texts, productive and interpretive processes, the relationship between texts, processes and
their conditions needs to be taken into account. The following figure can be seen as an
illustration of this relationship.
9
Figure 1: Discourse as text, interaction and context
(Fairclough, 2001:21)
However, CDA is made distinguished in terms of “critical”. According to Rogers, R.
(2004), the concept of critical is rooted in the Frankfurt school of critical theory (Adorno,
1973; Adorno & Horkeimer,1972; Habermas, 1976). Critical research and theory is a
rejection of naturalism (that social practices, labels, and programs represent reality),

matter of the conventions of discourse types constraining participants’ contributions in
these three ways – they may have long-term structural effects on an institutions or society.
(Figure 2. Fairclough, 2001,62).
Constraints Structural effects
Contents
Relations
Subjects
Knowledge and beliefs
Social relationships
Social identities
Table 1: Constraints on discourse and structural effects
For CDA, language is not powerful on its own. In fact, it gains power by the use powerful
people make of it. This explains why CDA often chooses the perspective of those who
suffer, and crtically analyses the language use of those in power. “CDA takes the
experiences and opinions of members of such groups seriously, and supports their struggle
against inequality. That is, CDA research combines what perhaps somewhat pompously
used to be called ‘solidarity with the oppressed’ with an attitude of opposition and dissent
against those who abuse text and talk in order to establish, confirm or legitimate their
abuse of power. Unlike much other scholarship, CDA does not deny but explicitly defines
and defends its own sociopolitical position. That is, CDA is biased – and proud of it”
(Teun van Dijk, Multidisciplinary CDA: a plea for diversity collected by Ruth Wodak &
Michael Meyer (Eds.), 2001: 96).
Fairclough and Wodak (1997) offered eight foundational principles of CDA. These
principles are a useful starting point for researchers interested in conducting CDA. They
include:
• CDA addresses social problems
• Power relations are discursive
• Discourse constitutes society and culture
11
• Discourse does ideological work

In the surveys, focus groups should also be identified and thoroughly analyzed. The
purpose of such groups at this stage is twofold: The researcher wants to know to what
extent the identification of specific scenes, media, and actions have reliability and validity
for members of the group under study, and they wants to understand how important or
salient the categories which have been identified are for the population being studied as
well. (Scollon, 2001: 158)
In terms of textual analysis, CDA strongly relies on linguistic categories. This does not
mean that topics play no role at all, but that the core operationalizations depend on
linguistic concepts such as actors, mode, time, tense, argumentation, etc. Van Dijk suggests
that the analysis should concentrate upon the following linguistic markers: Stress and
intonation, word order, lexical style, coherence, local semantic moves, topic choice,
speech acts, schematic organization, rhetorical figures, syntactic structures, turn takings,
repairs, hesitation.
Norman Fairclough in his book Language and Power (2001) provides a more detailed and
clearer method. In his opinion, “The systemic-functional theory of language is particularly
helpful in textual analysis (Halliday 1978; 1985; Hodge and Kress 1988; Thibault 1991),
both because its approach to studying grammar and other aspects of language form is a
functional one, and because it is systematically orientated to studying the relationship
between the texture of texts and their social contexts” and “Systemic-functional linguistics
also has a view of texts which is a potentially powerful basis not only for analysis of what
is in texts, but also for analysis of what is absent or omitted from texts.” (Fairclough,
Linguistic and intertextual analysis within discourse analysis, 1992) As a result, Fairclough
suggests first of all structural analysis of the context, and secondly interactional analysis,
which focuses on linguistic features such as: agents, time, tense, modality, and syntax.
More concretely, he sets out three stages of CDA as follows.
Three stages of CDA are description, interpretation and explanation. Description is the
stage which is concerned with the formal properties of the text. Interpretation is concerned
with the relationship between text and interaction – with seeing the text as a product of a
process of production, and as a resource in the process of interpretation. And explanation is
concerned with the relationship between interaction and social context – with the social

14
Are there important features of relational modality?
Are the pronouns we and you used and if so, how?
7. What expressive values do grammatical features have?
Are there important features of expressive modality?
8. How are (simple) sentences linked together?
What logical connectors are used?
Are complex sentences characterized by coordination or/ subordination?
What means are used for referring inside and outside the text?
C. Textual structures
9. What interactional conventions are used?
Are there ways in which one participant controls the turns of others?
10. What larger scale structures does the text have?
Fairclough (2001: 92-3)
The definitions of three terms: experiential, relational and expressive, are of great
importance to the understanding of the framework. In Language and Power (2001),
Fairclough claimed that formal features of texts have experiential, relational, expressive or
connective value, or some combination of these. By looking at experiential values CDA
attempts to show how ‘the text producer’s experience of the natural or social world’ (ibid:
93) effects and is shown in a text. A person’s views of the world can be identified by
assessing formal features with experiential value. Relational values may identify the
perceived social relationship between the producer of the text and its recipient. The third
dimension, expressive value, provides an insight into ‘the producer’s evaluation (in the
widest sense) of the bit of the reality it relates to.’ (ibid: 93) This should identify the
relevant parties to the text’s social identities. Fairclough (2002: 93) goes on to identify
another value that any formal feature may possess, connective value, as its function may be
to connect together parts of a text. He also stresses that ‘any given formal feature may
simultaneously have two or three of these values’ (ibid: 93). However, in my opinion,
Fairclough’s list of questions seems capable of generating an astonishing amount of
analysis and it may be less suitable for larger quantities of text.

16
Figure 2: Interpretation (Fairclough, 2001:119)
Figure 3: Explanation (Fairclough, 2001:136)
17
Social orders Situational context
Interactional history Intertextual context
Phonology, grammar,
vocabulary Source of utterance
Semantics,
Pragmatics Meaning of utterance
Cohesion,
Pragmatics Local coherence
Schemata Text structure and point
Societal Societal
Institutional MR Discourse MR Institutional
Situational Situational
Determinants Effects
1.2. Systemic Functional Linguistics and its role in CDA.
As Fairclough’s model is chosen as a base for my study, his method and analyzing
procedure have been presented in detail. Besides Fairclough’s model, Halliday’s Systemic
Functional Linguistics (SFL) is also taken into account and the following part is a brief
look at SFL and the reason why and how it is used in CDA.
SFL says that we perform functions through language, i.e. what we intend to do with a
piece of language. Clearly, speakers have reasons for saying something and for saying it in
the way they do. As a result, speakers have to make choices. SFL sets out to investigate
what the range of relevant choices are, both in the kinds of meanings that we might want to
express (or functions that we might want to perform) and in the kinds of wording that we
use to express these meanings, and to match these two sets of choices.
However, the term ‘choice’ does not necessarily imply a conscious process of selection by
the speaker, what SFL aims to uncover a functional analysis are the reasons why the

process is a kind of activity in people’s mind, requires a conscious participant such as
thinking, loving, wanting, hearing. Verbal process is the process of saying such as saying,
telling, speaking, talking. Relational process is the process of ‘being’, ‘having’ and ‘being
at’ in form of three subtypes: the intensive, the possessive, and the circumstantial.
Existential process is the process of existing, indicating that something or some natural
force exists. Table 2 is an overview of these processes.
Process types Category
meanings
Participants Example
Material :
Action
Event
‘doing’
‘doing’

‘happening’
Actor, Goal, Recipient
The mayor dissolved the
committee.
The mayor resigned.
Behavioural : ‘behaving’ Behaver, (Phenomenon) She cried softly.
Mental:
Perception
Affection
Cognition
‘feeling’
‘sensing’
‘emotive’
‘thinking’
Sensor, Phenomenon

simultaneously assigns a corresponding role, such as ‘informant’, to the other person.
Halliday (1994) provides a table to characterize the primary speech roles as follows.
20

modality
personal
evaluation
Interpersonal

enacted roles (speech roles)
interactive
projected roles
Commodity
exchange
Role in exchange
(a) goods -&- services (b) information
(i) giving Offer
Would you like this teapot?
Statement
He is giving her the
teapot.
(ii) demanding Command
Give me that teapot!
Question
What is he giving her?
Table 3: Primary speech roles (Halliday, 1994:69)
However, the speaker may also project a role on to himself/herself or the other person by
the way he/she talks about them. This is clearest with naming, where the way that the
speaker names the other person indicates how he/she thinks of that other person. For
example: Reader, I married him.

A theme is single when the thematic element itself is presented by just one constituent – a
nominal group, an adverbial group, or a prepositional phrase, or even a clause in the case
of predicated theme. Meanwhile, the theme is multiple when it has further internal
structure of its own. The following is the summary of components of multiple theme.
Metafunction Components of theme Example
Experiential
theme
Topical elements (participant, circumstances,
process)
Elements playing as
Actor / Agent , Goal /
Medium, Circumstance
in the clause
Textual
theme
Continuative elements
Structural elements (conjunction or WH-
relative)
Conjunctive elements (Adjunct)
Yes, no, well…
And, but…
Also, therefore…
Interpersonal
theme
Modal (adjunct)
Finite (operator)
WH – (interrogative)
Vocative element
Surely, maybe…
Don’t, would…

2. Specifying or changing the framework for the interpretation of the following clause
(or clauses). This is mostly done by the choice of marked Theme, especially
Adjunct, or a thematic equative or predicated Theme. A ‘heavy’ Subject theme,
giving a large amount of information, can also be used for this purpose.
23
3. Signalling the boundaries of sections in the text. This is often done by changing
from one type of Theme choice to another.
4. Signalling what the speaker thinks is a viable/ useful / important starting point. This
is done by repeatedly choosing the same element to appear in Theme (a particular
participant, the speaker’s evaluation, elements which signal interaction with the
hearer, etc.)
The above-mentioned functions of language have been proved to be useful in many
discourse analyses. In addition, there is one aspect that I found of particular importance to
my analysis of the text. It is the macrostructure of text. In Halliday ‘s viewpoint, the
macrostructure represents relations between blocks of sentences and the global
organization of texts, while the microstructure represents the relations between sequences
in actual text. The macrostructure of a text can be understood as the construction of global
organizational patterns. As communicative purpose plays an important role in determining
the macrostructure for writers/ speakers, when doing CDA, one cannot ignore mentioning
how ideology is revealed in the writer/ speaker’s choice of the text’s overall scheme.
24
Chapter 2:
A CRITICAL DISCOURSE ANALYSIS
OF MARTIN LUTHER KING’S SPEECH:
‘Beyond Vietnam – A Time to Break Silence’.
2.1. Textual description and Analysis
2.1.1. Analysis in terms of vocabulary
Text analysis is an indispensable process in discourse analysis generally and in CDA
particularly. As given by Faiclough (1989), the first stage in text analysis is description to
find out linguistic features such as features of vocabulary, grammar, types of speech act,