Annual Review of Applied Linguistics (1990) 11, 84-99- Printed in the USA.

Copyright © 1991 Cambridge University Press 0267-1905/91 $5.00 + .00

CRITICAL DISCOURSE ANALYSIS*

Gunther Kress

INTRODUCTION

The label Critical Discourse Analysis (CDA) is used by a significant number

of scholars with a diverse set of concerns in a number of disciplines. It is well-
exemplified by the editorial statement of the new journal Discourse and Society, which
defines its envisaged domain of enquiry as follows: "the reproduction of sexism and
racism through discourse; the legitimation of power; the manufacture of consent; the
role of politics, education and the media; the discursive reproduction of dominance
relation between groups; the imbalances in international communication and
information." While some practitioners of Critical Discourse Analysis might want to
amend this list here or there, the set of concerns sketched here well describes the field
of CDA. The only comment I would make, a comment crucial for many practitioners
of CDA, is to insist that these phenomena are to be found in the most unremarkable
and everyday of texts—and not only in texts which declare their special status in some
way. This scope, and the overtly political agenda, serves to set CDA off on the one
hand from other kinds of discourse analysis, and from textlinguistics (as well as from
pragmatics and sociolinguistics) on the other.

All forms of discourse analysis take texts to be the proper domain of linguis-
tic theory and description (rather than a focus on constituents of texts); all share an
interest in the understanding of extended text socially or at least contextually situated,
and in producing accounts of texts that draw on features of the context (social,
cultural, co-textual) to provide explanatory categories for the description of textual
characteristics. Conversely, most forms of discourse analysis aim to provide a better
understanding of socio-cultural aspects of texts, via socially situated accounts of texts.

Critical Discourse Analysis shares these aims of discourse analysis. Indeed,

like textlinguistics, it aims to provide accounts of the production, internal structure, and
overall organization of texts. Where it differs crucially from both is precisely in its
aims to provide a critical dimension in its theoretical and descriptive accounts of texts.

84
 CRITICAL DISCOURSE ANALYSIS 85

Implicitly or explicitly, practitioners of CDA have the larger political aim of putting
the forms of texts, the processes of production of texts, and the process of reading,
together with the structures of power which have given rise to them, into crisis. By
denaturalizing the discursive practices and the texts of a society, treated as a set of
discursively linked communities, and by making visible and apparent that which may
previously have been invisible and seemingly natural, they intend to show the imbrica-
tion of linguistic-discursive practices with the wider socio-political structures of power
and domination. In as far as these structures act to the detriment of particular groups
in a society, critical discourse analysts hope to bring about change not only to the
discursive practices, but also to the socio-political practices and structures supporting
the discursive practices.

Given the prevailing ethos (CDA practitioners would say "ideology") of

"value-free" science, of objectivity, and of the political/ideological neutrality of
scientific practice, CDA is an openly political and therefore potentially contentious
activity. Practitioners of CDA see no problem in acknowledging this. They would
point to the long history of debate on the itself contentious question of a value-free,
and therefore in a fundamental sense a-historical, science—particularly in the era of
postmodernism in which certain forms of discourse analysis (e.g., narratology) have
provided searching critiques of the foundations and constitution of disciplinary
knowledge. They would insist that, while their activity is politically committed, it is
nonetheless properly scientific, perhaps all the more so for being aware of its own
political, ideological, and ethical stance.

This does mean, however, that applying the label of Critical Discourse
Analysis to a particular project is to place it in a potentially contentious domain. I
have therefore been relatively circumspect in applying this label, in not attempting to
'appropriate' work to the paradigm of CDA. There is, however, much work in
sociolinguistics, discourse analysis, and forms of pragmatics that has both the effect
and frequently the intention of bringing about changes in social practices through
modifications—more or less far-reaching—of linguistic-discursive practices. One
would cite here, merely as exemplification, work on professional-client interactions,
court-room language, gender-based differences in language form and use, national
language questions, literacy, multi-cultural/multi-lingual policies, etc.

THEORETICAL FOUNDATIONS AND SOURCES

There are certain criteria characterizing work in the Critical Discourse

Analysis paradigm which do distinguish it clearly from other politically engaged
discourse analysis work. Together with the domain sketched above, and assumptions
shared with discourse analysis and textlinguistics, they provide an integrated theoretical
framework for CDA. They are listed here in outline only with minimal elaboration;
further discussion is provided in the readings. I should add that this outline formalizes
what is to some extent left implicit in much CDA writing.

1. Language is first and foremost a type of social practice. It is one among many
social practices of representation and signification, along with visual images,
86 GUNTHER KRESS

music, layout, gesture, etc. In the case of Western societies it is the most fully
described, the most heavily theorized, and possibly the most fully articulated
semiotic system.

2. Texts are the result of the actions of socially situated speakers and writers; these
speakers and writers operate with relative degrees of possibilities of choice always
within structurings of power/domination.

3. The relations of participants in the production of texts are generally unequal,

ranging from a hypothetical state of complete equality (the dimension of
solidarity) to complete inequality (the dimension of power).

4. Meanings are the result of the (inter)action of readers and hearers with texts and
with the speaker/writers of texts; they are always subject to more or less closely
enforced normative rules (for instance, rules of genre), and to relations of power
obtaining in this interaction.

5. Linguistic features (as signs) at any level are the result of social processes, and
hence are motivated conjunctions of forms (signifiers) and meanings (signifieds).
Thus, linguistic features are never arbitrary conjuncts of form and meaning.

6. Linguistic features, at any level, in their occurrence in texts are always

characterized by opacity; language itself is an opaque medium and so, conse-
quently, are texts.

7. Users of language, due to their socio-cultural positioning, have a particular stance

towards the set of codes which make up a language. That is, language users, as
socially located individuals, do not have access to 'the language system' as such,
but have partial and selective/selected access to particular configurations of the
language system. This means that in most interactions the producers of a text
bring to bear different dispositions towards language, different knowledge of
differently configured systems, and different know-ledge of textual forms. These
differences are closely related to and enmeshed with the differing social
positionings of language users.

8. Points 1 through 7 argue that the notion of 'the language system' is a highly
problematic one in CDA, as are also notions such as 'norm,' 'core,' etc.

9. Points 1 through 8 necessitate a view of language that always and everywhere

takes history (as ideologically/politically inflected time) into account, whether it be
the micro-history of a spoken interaction (conversation, interview, discussion) or
the larger histories of social/linguistic institutions.

10. Last, but not least, CDA must rely ultimately always on quite precise analyses and
descriptions of the materiality of language—on close linguistic description. This
too is a criterial characteristic of CDA, and marks it off from other kinds of
discourse analysis.
 CRITICAL DISCOURSE ANALYSIS 87

Before passing on to a brief account that ties this characterization into a more
specific history of theoretical and descriptive work in CDA, a few additional comments
may be helpful. Clearly, the 'angle of attack' of CDA is fundamentally different from
forms of textual analysis founded on the notion of an autonomous linguistic system.
In CDA the idea of autonomy makes no sense, and the notion of the linguistic system
is, as I have mentioned, highly problematic. CDA works from the social to the
linguistic, or better, sees the linguistic as within the social. So whereas pragmatics
functions, in a sense, to deal with problems of 'use' exported from an autonomously
conceived linguistics to a marginal discipline, such questions are entirely within the
ordinary domain of CDA. Or while certain kinds of sociolinguistics function to
establish correlations between autonomously theorized linguistic domains and the social
context (e.g., a linguistic feature correlates with a social grouping), CDA treats that
relation as one of (more or less strong) determination between differing domains of the
social situation.

It should be stressed that while the definition of the domain and the 'angle of
attack' is thoroughly social, this does not preclude attention to (social) psychological
and cognitive dimensions of discourse and discourse processing. Nevertheless, the
psychological make-up—the constitution of individuals as 'social subjects'—is seen as
an aspect of the social, as are cognitive structures. That is, speaker/hearers or writ-
ers/readers have particular subjectivities due to the characteristics of the social
structures and processes which formed their histories, and in which they are located.
The 'critical' aspect of CDA, what is to be put into crisis, is the social, not the
psychological (other than as the effect of the social). In other words, CDA far from
ruling out social psychological or cognitive aspects of discourse and discourse process-
ing, treats these as integral aspects of the project.

In fact, an understanding of the social/psychological makeup of individuals—a

theorization of social subjects and of subjectivities—is crucial in CDA. On the one
hand, CDA attempts to describe structurings of power and domination, their reproduc-
tion in and through texts, and their effects on the possibilities of individual action. On
the other hand, CDA practitioners wish to point out precisely what possibilities of real
freedom of action—that is, not fully determined, not fully constrained—is open to the
individual speaker/writer or hearer/reader. CDA does not paint a picture of unremit-
ting determinism. Theoretically, it cannot do so; the differential placing of social
individuals (or to use a term coined by Basil Bernstein, their differing 'coding-
orientations') inevitably leads to degrees of openness, and to relatively unpredictable
outcomes, in discursive interactions.

In this connection, the concept of choice is an essential one. In a functional

grammar, such as Halliday's (1985), language is theorized as consisting of systems of
linguistic features linked in networks mapping both optional and obligatory co-
selections of features. So, for example, at the level of a clause, a speaker must select
from among the (mood) features 'interrogative,' 'declarative,' 'imperative'; if the
selection 'interrogative' or 'declarative' is made, the speaker must select from the
system of tense. The requirement to make a selection is obligatory; the kind of
selection made is relatively less so. "Can I take the car this afternoon?" can in certain
88 GUNTHER KRESS

circumstance be an optional alternative to "I can take the car this afternoon, can't I?"
While each of these forms projects a somewhat different social relationship between
speaker and hearer, there are nevertheless circumstances where either one or other of
the two forms can be used by the speaker. That is, the speaker has the possibility of
real choice, a choice to project, and therefore make the attempt to construct one rather
than another social relation.

There are also situations where this choice does not exist, where a greater
degree of power difference between speaker and hearer rules this out. In that situation
there might nevertheless exist the possibility of a choice of different tense (rather than
different mood): "Can I take the car this afternoon?" rather than "Could I take the car
this afternoon?" (a choice in the system of modality). Again, each of these two forms
projects a specific social relationship. There will be situations where no choice exists,
situations of overwhelming power-difference, or highly regulated situations (e.g.,
rituals), but they are likely to be few. Most often, linguistic interactions are likely to
be those in which the degree of choice and its characteristics are the issue, rather than
absence of the possibility of choice. 'Choice' is the category that captures and
reflects, on the one hand, degrees of power and control at issue in an interaction, and
on the other, the potential degrees and characteristics of real—not determinate—action
which are available to participants in linguistic interactions, whether spoken or written.

Concerns of this kind, of power, of the possibilities of individual action, and

of critique (or to use the fashionable term—deconstruction) are contemporary concerns,
and have been so—in the current phase—since the early seventies. Indeed, one major
component of CDA, Critical Linguistics, quite self-consciously adapted its label from
"critical sociology." Critical Linguistics, a term first adopted by a group of scholars
working at the University of East Anglia (Fowler, et al. 1979) had as its two aims: 1)
to use the tools provided by linguistic theories (in this case both a 1965 version of
Chomskyan grammar, and Halliday's theory of Systemic Functional Linguistics) to
uncover the linguistic structures of power in texts, and 2) to make the discipline of
linguistics itself more accountable, more responsible, and more responsive to questions
of social equity. (For accounts see Fairclough 1989a, Fowler 1987; 1990, Fowler, et al.
1979, Hodge and Kress 1974; 1988, Kress 1989, Kress and Hodge 1979.) In its use of
descriptive/diagnostic tools, Critical Linguistics has been eclectic; as mentioned above,
in early versions realist interpretations of transformations were integrated into an
overarching functional grammar framework (as developed by Halliday). This eclecti-
cism—which continues to be a characteristic of CDA—has been defended on two
grounds: 1) it is unwise to neglect linguistic insights produced by generations of
scholars, as well as by current work in linguistics; 2) the eclecticism does not operate
at the level of the theoretical framework, where there continues to be efforts to
develop a coherent theoretical approach.

Critical Linguistics has from the first taken text as the relevant linguistic unit,
both in theory and in description/analysis. However, its analyses also drew largely on
analytical descriptive categories from sentence and below-sentence grammar. Catego-
ries that have been particularly prominent have been transformations, transitivity types
(or case-grammar analyses), modality forms (modal auxiliaries, adverbial modifiers,
 CRITICAL DISCOURSE ANALYSIS 89

mental process verbs), forms of embedding, and subordination and coordination. From
the functional grammar of Halliday (Halliday and Hasan 1976) has come an interest in
devices of textual structuring, particularly cohesive devices of various kinds.
Halliday's work on the intonational/informational structure of spoken language has
been developed into investigations into systematic differences among spoken and
written varieties of language (Kress 1982; 1989, Halliday 1989), and the social and
political effects and uses of this differentiation.

The analysis of the textual dimensions of larger linguistic units in CDA has
drawn on a number of sources, notably the extensive work of, among others, van Dijk
(van Dijk and Kintsch 1983). His examination of macro-structures of text and smaller
subsidiary micro-structures within these, organized as schemata, have proven to be
important diagnostic devices in the analysis of all kinds of textual form. This has been
complemented by analyses of the textual structures of spoken interactions, as in
classroom discourse, doctor-patient interviews, etc. (particularly in work done at the
University of Lancaster). Some practitioners within CDA have drawn quite extensively
on certain resources of pragmatics (Chilton 1985, Fairclough 1989a), including speech
act analysis, conversational implicature, and presuppositional structures. They have
also made use of certain analyses of cognitive structures (Chilton 1985, van Dijk
1988). The analysis of metaphor and metaphoric structures has also been a prominent
element of CDA since the earliest work in Critical Linguistics (Chilton 1985,
Fairclough 1988, Fowler, et al. 1979, Halliday 1985, Kress 1989). CDA consequently
has a wide range of descriptive/methodological categories; in the practices of CDA,
each finds a place and use within a particular theoretical framework.

THE PROJECT OF CRITICAL DISCOURSE ANALYSIS IN PRACTICE

In the small example discussed above it is apparent that language is doing

more than merely representing a clear-cut state of affairs (e.g., Joan needs a car; Joan
doesn't have a car; Ann has a car; Joan requests the loan of Ann's car). There quite
clearly is an aspect of representation, but there is also a considerable element of
construction, or projection. That is, the state of affairs to be represented is produced
in such a way that it either reflects, more or less closely, the social relations of the
participants, or it projects a particular—more or less plausible—version of such a
relation. The motivating dynamic for the choice of one form rather than another is
power-difference. Joan and Ann are not equal in power in this situation: Ann has
more power. Of course it could be that Joan is Ann's superior at work, and will soon
be writing a report on Ann in relation to a promotion she desperately wants, of which
Ann is well aware. Joan might then be seen to have more power, a fact which
immediately affects her possibilities of (linguistic/social) choice. Joan can use this
greater range of choice in different ways; she might be extremely polite, disguising
power: "Look Ann, I'm awfully sorry, but could I possibly trouble you to have just a
little loan of your car this afternoon?" She might choose to assert this power: "Ann,
I'm sure you don't mind if I borrow your car this afternoon, do you?" Choice enables
the possibility of changing the actual, or projected social relations and structures; the
possibility of choice comes with greater power. Being linguistically, or semiotically
more specific, one effect of power is a move away from relatively straightforward
 90 GUNTHER KRESS

representation—power deforms communication. The above example, however, suggests

the possibility of undeformed communication, and therefore needs to be amended
somewhat. Given that the available linguistic forms were formed by past interactions,
in each of which power-difference played its part, no linguistic form is itself neutral;
rather, all linguistic forms always express a particular stance, modality, or inflection of
• the thing to be represented. In other words, the materiality of language itself has
meaning, so that to use language is always, inevitably, to enter into particular posi-
tions.

The commonsense view of language assumes that it represents, more or less

! felicitously, some event or state of affairs in the world. Mainstream linguistics as well
| as many kinds of discourse analysis hold views very close to this: neither the notion
of mediation, nor the notion of a systematic distortion of communication through the
effects of power, form a part of linguistic theory. This represents a serious limitation
on the explanatory power of such theories. A quite fundamental aspect of linguistic
(interaction, then, cannot appear in the agenda of concerns, let alone find possible
answers. First, in its representational function, language form is always deformed by
the effects of power, leading to mystification; second, in its representational function,
the material form of language always has a mediating effect, necessarily leading to
skew in certain ways; and third, in its constructing function, language always projects
social relations and structures as wished by the participants, normally those of the
more powerful participant There are thus at least three quite distinct characteristics
through which texts are shifted away from straightforward representation, and towards
mystification. Each has far-reaching effects in communication, and effects on the
forms of language itself. All linguistic forms and processes are affected at every level:
from the phonological, via the syntactic/grammatical, to the textual. Hence no aspect
of form can, theoretically, be ignored in analyses and descriptions provided by CD A,
whether it be rhetorical devices or strategies, text-creating devices and formal aspects
of textual structure, or the more 'expected' syntactical categories and forms.

Here I will cite four examples briefly, simply to indicate the kinds of forms
involved: the textual category of genre, the rhetorical strategy of presupposition, the
lexical category of lexicalization, and the syntactic category of word-order.

1. Genre In CDA, genre is a category which explains conventionalized and

conventionally available textual forms not, as is usually the case, in terms of
reified historical/linguistic categories, but rather in terms of the contingent
structurings of social occasions, the organization of social participants, and
their purposes and intentions. Hence, genres are always seen as the linguistic
products of particular social occasions, encoding the social organization,
structures, etc. of that occasion. A question about generic form is thus never
answerable simply by reference to some transcendent formal categories: 'the
epic,' 'the sonnet,' or now, 'the interview,' 'the editorial.' Rather such a
question leads to answers involving a description of a social occasion, its
place and value in the larger social structures, and an assessment of who can
or cannot participate, and in what ways. Textual form thus becomes a matter
of the greatest interest, not simply in itself, but as a means for gaining access
 CRITICAL DISCOURSE ANALYSIS 91

to an understanding of social and cultural organization. Genres tend to

present themselves—or tend to be presented—as uncontentious,
uncontroversial, naturalized if not natural forms; interviews are what they are
because that's the most efficient way of conducting business. This view, of
course, leaves unasked the most important questions, not the least being one
about the very existence—a very recent one in the case of the interview—of a
genre form itself.

2. The rhetorical strategy of statement via a presuppositional structure This has

been widely studied in pragmatics; for instance, it involves the 'smuggling in'
of contentious material in a manner in which it is difficult to contest Here I
want to draw attention to the use of intonation features to achieve this in
speech. Assume someone says "I hear you haven't registered to pay the
community charge yet"; it will make a considerable difference whether the
speaker makes you or community intonationally prominent, or perhaps both.
Consider what social relations between speaker and hearer are projected by
any of the three forms (not said in jest), and therefore, who is likely to use
which form?

3. Lexicalization Lexicalization touches on the power of language not just to

represent and mediate, but to project a reality which exists only as the effect
of certain resources of language—and the requisite power to use these forms.
To take a trivial example: if I stay with a well-to-do friend, and he asks me to
take off my ordinary shoes to put on a pair of pool-side shoes, I'm then likely
to be somewhat nervous but unsurprised if he should offer me a pair of
inside-shoes (TV-shoes perhaps?) when we leave the pool. In other words,
the productivity of language in this domain enables the possibility of a large,
perhaps infinite, set of verbal categories for which physical objects have to be
constructed, or at least imagined. (I might ask my friend, timidly pointing to
my outside-shoes: "are these OK as TV-shoes?") Less trivial examples are
known to all of us, as when institutions produce lists of rules, regulations,
offences, penalties, etc. The effect relies on the notion of language as a set of
names for objects. Here language mystifies by producing labels for objects
yet to be invented, and naturalizes these by relying on the ingrained, common-
sensical notions of the referentiality of language. We all know this as a
bureaucratic means of exercising control. Closely akin as strategies are so-
called euphemistic labellings, as when the US Pentagon called one of its
nuclear ballistic missiles "The Peacemaker.'

4. Word order We know that different conceptions of a relationship are in-

volved in saying Mary married Bill, or Bill married Mary, or Bill and Mary
married, etc. The triviality of this example allows us to realize that all
aspects of form are always subject to the effects of social forces. Wide-
reaching critiques of gender-structurings mean that the second of these
examples is now highly contentious, likely to be taken as offensive. Should
this socio-cultural situation persist, a particular word-order, in relation to this
92 GUNTHER KRESS

domain, may become obligatory; social factors will be exercising their effects
on grammar.

These examples are not illustrating matters that have not been recognized,
though the kind of account provided here is ruled out in many theoretical approaches.
What is different, decisively, is that in CDA all aspects of textual/linguistic form are
analyzed, described, and accounted for from within a framework of socio-cultural
practice. There are, consequently, not linguistic explanations on the one hand, textual
ones on the other, and occasional, isolated and theoretically unconnected correlations
with aspects of the socio-cultural situation. In CDA the analyses are necessarily
embedded in a socio-cultural theory of communication. This means that analyses of
texts take into account all relevant features of the socio-cultural situation as a commu-
nicative situation; including the position, role, and subjectivity of the hearer/reader, and
the medium/channel. An analysis of text is thus always a (partial) analysis and
description of aspects of the socio-cultural situation.

Within this framework, CDA has examined language in the mass-media, in

education, in advertising, in objects of popular culture, in professional settings
(professional languages, and professional-client interactions), in bureaucracies and in
the bureaucratization and technologization of language, in political contexts whether
nationally or internationally, in objects of high culture (literature particularly), and in
public discourses such as law, medicine, science, economics, and the economy. Issues
dealt with have included racism; gender structures; literacy and questions of education-
al disadvantage; ethnicity and multi-ethnic situations; class, age, professionalism, and
their effects on language; power and control in their various manifestations; and
ideology and its appearance through text

It may be worth pointing to an important characteristic of CDA in comparison

with other forms of discourse analysis. CDA is not interested, in a crucial sense, in
language and advertising, even less so in the language of advertising (or politics, or the
media, or education). These formulations and their projects assume a discreteness of
the language of advertising from the language of education (or whatever), as though
one was dealing with quite different domains. CDA, quite to the contrary, would
emphasize the continuity between language in advertising and language in education,
for instance, in an attempt to focus on the very real areas of socio-cultural and
linguistic/textual continuity between these domains—and with all others. This, of
course, has very real advantages in work across fields of application, (e.g., from
education to ESL, and to stylistics).

CRITICAL DISCOURSE ANALYSIS AND ITS APPLICATIONS

CDA is concerned to reveal the complex of processes involved in the

production, communication, and reception/reproduction of texts. Because it always
takes texts that occur regularly in important domains of socio-cultural life, the results
which it provides always have (at least the possibility of) direct application. If non-
English-speaking learners of English can be taught the entire connection between
power-structures and politeness-conventions, they will find it relatively easy to
 CRITICAL DISCOURSE ANALYSIS 93

comprehend the use of linguistic distancing devices as markers in expressions of

politeness. If, further, there is a theory which insists that signs are always iconic, that
is, motivated conjunctions of signifiers and signified, metaphors in fact, it will be
relatively easy to show why past tense (distance or remoteness in time) can serve as a
metaphor for the expression of social distance or remoteness: "I want to ask you, can I
bother you for a moment" contrasted with the 'more polite' "I wanted to ask you,
can/could I bother you..." If on the other hand one is interested in developing an adult
literacy program, it is necessary to know what forms of texts (genres) are most
powerful in relation to the group for which the program is being developed, as well as
knowing what syntactic/grammatical forms are most characteristic of certain kinds of
texts. But more than that, students will need to know what socio-cultural meanings are
coded in particular (aspects of) genres so that they may gain the facility to produce the
literate forms which they deem most useful for their purposes.

Any application of discourse analysis is bound to be interested in an under-

standing of the production and reproduction (reading/hearing) of whole texts. A theory
which situates its accounts in the socio-cultural is likely to provide not only formally
apt, but culturally meaningful descriptions. In fact, it is likely to provide analyses that
attempt to describe the categories which are generative in the production of discourse.

It is for this reason that CDA has applications not only in the traditional
domains of applied linguistics, but applications in other domains and disciplines. If
texts are repositories of the effects of socio-cultural practices, then an adequate theory
of text should turn texts into useful resources for socio-cultural analysis—whether in
history, sociology, cultural studies, or elsewhere. It is also for this reason that
practitioners in a range of disciplines other than applied linguistics have found CDA
particularly useful; it allows for readings of texts that go well beyond the superficially
accessible meanings and that provide another means of gaining access to socio-cultural
organization.

PROBLEMS AND DEVELOPMENTS

Because of its concern with texts from domains that are always in some sense
problematical, CDA is already a kind of applied linguistics. It uses linguistic
methodologies to address problems of the widest kinds, and in doing so provides
essential means for solutions. To become more successful in its own terms, CDA
needs to address the major problem facing it at the moment. That concern is the
relative dispersal of effort, and a consequent under-development of theoretical support.
While the picture which I have presented would, I am confident, find wide assent,
there has been relatively little overt effort to address theoretical issues facing CDA.
Rather, practitioners have felt the need to bring their theoretical and descriptive skills
to bear on always urgent problems: disadvantage, oppression, and loss of opportunity
in so many fields. Hence the account of CDA that I have given here is an optimistic
reading of actual achievement and of perceived potential. So much more could be
achieved with a somewhat greater degree of coordination and theoretical dialogue
among practitioners.
94 GUNTHER KRESS

In the meantime, even in the absence of such coordination, CDA is moving

beyond the twin aims of putting social/linguistic structures into crisis, through use of
the devices provided by 'mainstream linguistics,' and making the discipline of
linguistics itself more responsive and responsible to pressing social and political issues.
In a very real sense, CDA is emerging as a distinct theory of language, a radically
different kind of linguistics. The characteristics which I outlined at the beginning
strongly suggest just that. A whole range of work is taking place within this broadly
sketched framework. Far-reaching theoretical repositionings have already taken place,
such as the placing of history in the very center, the denial of the stable linguistic
system, and an installing of language users as socially located individuals into the
center as agents in the processes of linguistic/textual production and of linguistic/social
change. In Australian work, the taken-for-grantedness of language at the centre of
culture and communication is coming under sustained questioning in the strand of
CDA known as 'Social Semiotics.' Here real efforts are made to understand systems
of representation other than language—visual images, music, and performance. This
understanding is then 'turned back' on language in new theorizations of the character-
istics of language.

NOTES

*I would like to thank Jill Brewster for her help in the preparation of this
chapter. It was written during a stay as a visiting Research Fellow in the Department
of Modern English Language and Linguistics at the University of Lancaster, whose
assistance I gratefully acknowledge.

ANNOTATED BIBLIOGRAPHY

Birch, D. 1989. Language, literature and critical practice. London: Routledge and
Kegan Paul.

This volume provides a comprehensive summary of major strands in CDA,

with a wide range of texts analyzed from the point of view of specific
methodological categories.

Birch, D. and M. O'Toole (eds.) 1988. Functions of style. London: Frances Pinter.

This collection provides detailed analyses and descriptions of a variety of

literary and non-literacy texts. It is useful in demonstrating the consistent
application of a range of methodological tools in CDA, and in illustrating
socially and culturally motivated accounts of style.
 CRITICAL DISCOURSE ANALYSIS 95

Chilton, P. 1985. Language and the nuclear arms debate: Nukespeak today. London:
Frances Pinter.

A consistent and sustained application of the theoretical framework of CDA is

applied to a set of texts around a particular overtly political issue. The
chapters develop analyses from specific theoretical/descriptive points of view
(metaphor, rhetorical strategies, causal syntax, lexis). The book is useful both
as an example of sustained and detailed analysis in CDA, as well as an
example of a more general exploration of the interrelations of politics and
language.

Dijk, T. A. van. 1987. News analysis. Case studies in national and international news
in the press: Lebanon, ethnic minorities, refugees and squatters. Hillsdale,
NJ: Lawrence Erlbaum.

Dijk, T. A. van. 1988. News as discourse. Hillsdale, NJ: Lawrence Erlbaum.

The two books together constitute a sustained and detailed exploration of

particular discourses, and their appearance and function in the media. The
second sets out the theoretical/descriptive approach; the first supplies rich
exemplifications.

Fairclough, N. 1989a. Language and Power. London: Longman.

This volume is a precise, carefully worked out accout of the social theories
underpinning CDA. At the same time, it is richly illustrated with a variety of
textual examples which are discussed with the use of a range of specific
linguistic/social categories. Both as a theoretical background, and handbook
of how to carry out CDA, it is highly useful.

Fowler, R., et al. 1979. Language and control. London: Routledge and Kegan Paul.

This book is a valuable early foundational text in Critical Linguistics. It

includes a large number of detailed critical linguistic analyses.

Halliday, M. A. K. 1985. An introduction to functional grammar. London: Edward

Arnold.

This comprehensive grammar constitutes an explicit account of a particular

kind of functional grammar, and does so through a detailed description of the
grammar of English. Both the theory and the description have been highly
influential in the development of Critical Linguistics and CDA, and are drawn
on in much CDA work.
96 GUNTHER KRESS

Hodge, R. 1989. Literature as communication. Cambridge: Polity Press.

This volume is a consistent application of the CDA/Critical Linguistics

framework to literary texts, leading both to a highly innovative approach to
literature and to assessments of particular literary texts.

Hodge, R. and G. Kress. 1988. Social semiotics. Cambridge: Polity Press.

The book starts from the assumption that no single code can be successfully
studied or fully understood in isolation because meaning resides so persua-
sively in a multiplicity of visual, aural, and behavioral codes. Consequently,
this book applies the theoretical framework of CDA and Critical Linguistics to
a wide range of semiotic media and forms: photographic images; TV; comics;
written and spoken language; sculpture; fashion; architecture; high and low
culture; and media, education, and advertising. From a multi-semiotic stand-
point, the volume provides a distinctive, critical perspective on these media
and on language itself.

Kress, G. 1989. Linguistic processes in sociocultural practice. London: Oxford Univer-

sity Press.

The book develops the social and cultural categories necessary to a full
understanding of the production and reception of texts. At the same time, it
provides a detailed linguistic analysis of a wide range of texts to show how
socio-cultural categories find specific linguistic expressions.

Kress, G. and R. Hodge. 1979. Language as ideology. London: Routledge and Kegan
Paul.

This book remains a valuable foundational text in Critical Linguistics. It

contains a large number of detailed critical linguistic analyses. It also sets out
the theoretical foundations of the first stage of Critical Linguistics.

Kress, G. and T. van Leeuwen. 1990. Reading images. Geelong, Victoria: Deakin
University Press.

Written from within the framework of CDA and Social Semiotics, this richly
illustrated book develops a systematic approach to the analysis of visual
images as a fully specific medium of representation. The book is particularly
written to allow teachers to understand the images produced by children as
part of the texts they produce, as well as providing a means of reading the
compositional structures (layout, images) of the texts children are asked to
read (e.g., textbooks).
 CRITICAL DISCOURSE ANALYSIS 97

Martin, J. R. 1989. Factual writing: Exploring and challenging social reality. London:
Oxford University Press.

This book is a meticulously detailed linguistic description of bureaucratic and

political writing in the framework of CDA.

Mey, J. L. 1985. Whose language. Amsterdam: John Benjamins.

Written by a leading practitioner of European pragmatics, this volume pro-

vides detailed accounts of a wide variety of texts in a framework highly
congenial to that of CDA.

Richardson, K. 1987. Critical linguistics and textual diagnosis. Text. 7.2.145-163.

This article represents a detailed and meticulous description of some of the

major critiques of CDA and Critical Linguistics. While very useful for that, it
is more so important for the theoretical issues it raises for CDA, and for
language theories generally.

UNANNOTATED BIBLIOGRAPHY

Aers, D. R., R. Hodge and G. R. Kress. 1982. Literature, language and society in
England 1580-1680. Dublin: Gill and Macmillan.
*Bourdieu, P. 1984. Distinction: A social critique of the judgment of taste. London:
Routledge and Kegan Paul. [Tr. R. Nice.]
Chilton, P. 1988. Orwellian language and the media. London: Pluto Press.
Clark, R., et al. 1987. Critical language awareness. Lancaster: Centre for language in
social life, University of Lancaster. fCLSL Working Papers 1.]
*Dews, P. 1987. The logics of disintegration: Post-structuralist thought and the claims
of critical theory. London: Verso.
Dijk, T.A. van and W. Kintsch. 1983. Strategies of discourse comprehension. London:
Academic Press.
Fairclough, N. 1988. Discourse in social change: A conflictual view. Lancaster: Centre
for Language in Social Life, University of Lancaster. Mimeo.
1989b. Language and ideology. In M. Knowles and K. Malmkjaer (eds.)
Language and ideology. Birmingham: English Language Research Centre:
University of Birmingham. 9-27.
1989c. What might we mean by 'enterprise discourse'? Lancaster:
Centre for Language in Social Life, University of Lancaster. [Research Paper
14.]
1990. Technologization of discourse. Lancaster: Centre for Language in
Social Life, University of Lancaster. Mimeo.
1991. Discourse and social change. Cambridge: Polity Press.
98 GUNTHER KRESS

Fairclough, N. In press. Critical linguistics. In W. Bright, et al. (eds.) Oxford

international encyclopedia of linguistics. New York: Oxford University Press.
*Foucault, M. 1972a. Orders of discourse. Social science information. 10.2.7-30.
1972b. The archeology of knowledge. London: Tavistock Publications.
[Tr. A. Sheridan Smith.]
Fowler, R. G. 1982. Linguistic criticism. London: Oxford University Press.
1987. Notes on critical linguistics. In R. Steele and T. Threadgold (eds.)
Language topics. Amsterdam: John Benjamins.
1990. Language in the news: Discourse and ideology in the press.
London: Routledge and Kegan Paul.
Forthcoming. Critical linguistics. In K. Malmkjaer (ed.) The linguistics
encyclopaedia. London: Routledge and Kegan Paul.
*Habermas, J. 1984. Theory of communicative action. Vol. I: Reason and the rational-
ization of society. London: Heineman. [Tr. T. McCarthy.]
Halliday, M. A. K. 1989. Spoken and written language. Oxford: Oxford University
Press.
and R. Hasan. 1976. Cohesion in English. London: Longman.
1989. Language, context and text: A social semiotic perspective.
Oxford: Oxford University Press.
Hasan, R. 1986. The ontogenesis of ideology: An interpretation of mother-child talk. In
T. Threadgold, et al. (eds.) Semiotics, ideology, language. Sydney: Sydney
Association for Studies in Society and Culture. 125-146. [Sydney Studies in
Society and Culture. No. 3.]
Hodge, R. and G. R. Kress. 1974. Transformations, models and processes: Towards a
useable linguistics. Journal of literary semantics. 4.1.4-18.
and D. Tripp. 1986. Children and television. Cambridge: Polity Press.
Kress, G. R. 1982. Learning to write. London: Routledge and Kegan Paul.
1991a. Towards a social theory of language. Cambridge: Polity Press.
1991b. Writing as social process. London: Routledge and Kegan Paul.
and T. Threadgold. 1988. Towards a social theory of genre. Southern
review. 21.3.215-243.
Luke, A. 1989. Open and closed texts: The semantic/ideological analysis of curricular
narratives. Journal of pragmatics. 13.1.
McHoul, A. W. and A. Luke (eds.) 1989. Discourse analysis in Australia, [special
issue of Journal of pragmatics. 13.2.]
Sinclair, J. and M. Coulthard. 1975. Towards an analysis of discourse. London: Oxford
University Press.
Steele, R. and T. Threadgold (eds.) 1988. Language topics: Essays in honour of
Michael Halliday. 2 Vols. Amsterdam: John Benjamins.
Thibault, P. 1986. Text, discourse and content: A social semiotic perspective. Toronto:
Toronto Semiotic Circle. [Toronto semiotic circle monographs. Vol. 3.]
1989a. Genres, codes and pedagogy: Towards a critical social semiotic
account. Southern review. 21.3.243-264.
1989b. Semantic variation, social heteroglossia, intertextuality: Thematic
and axiological meaning in spoken discourse. Critical studies. 1.2.181-209.
Threadgold, T. 1987. The semiotics of Halliday, Voloshinov and Eco. American
journal of semiotics. 4.3.107-142.
 CRITICAL DISCOURSE ANALYSIS 99

Threadgold, T. 1989. Talking about genre: Ideologies and incompatible discourses.

Journal of cultural studies. 3.3.208-29.
, et al. (eds.) 1986. Semiotics, ideology, language. Sydney: Sydney
Association for Studies in Society and Culture. [Sydney Studies in Society
and Culture. No. 3.]
van Leeuwen, T. 1985. Rhythmic structures of the film text In T. van Dijk (ed.)
Discourse and communication: New approaches Jo the analysis of mass
media, discourse and communication. Berlin: Walter de Gruyter.
1986. The consumer, the producer, and the state: Analysis of a
television news item. In T. Threadgold, et al. (eds.) Semiotics, ideology,
language. Sydney: Sydney Association for Studies in Society and Culture.
203-224. [Sydney Studies in Society and Culture. No. 3.]
1987a. Generic strategies in press journalism. Australian review of
applied linguistics. 10.2.199-220.
1987b. Semiotics of easy listening music: Changed times, changed
tunes. Sydney: Sydney Association for Studies in Society and Culture. Mimeo.
*Voloshinov, V. I. 1973. Marxism and the philosophy of language. New York:
Seminar Press.

This bibliography includes some sociological/philosophical items (*) important for the
theoretical background to CDA.