Discourse: A Critical Introduction available in Paperback
Organized thematically, this introduction outlines the basic principles and moves on to examine the methods and theory of CDA (critical discourse analysis). Topics covered include text and context, language and inequality, choice and determination, history and process, ideology and identity. Jan Blommaert focuses on how language can offer a crucial understanding of wider aspects of power relations, arguing that CDA should specifically analyze the effects of power.
About the Author
Read an Excerpt
Cambridge University Press
0521828171 - Discourse - A Critical Introduction - by Jan Blommaert
1.1 WHAT ARE WE TALKING ABOUT?
Power is not a bad thing - those who are in power will confirm it. They will argue convincingly that power is necessary in every system, for it is often that which allows the system to function in particular ways, without which the system would disintegrate or cease to operate effectively. Yet, power is a concern to many people, something that is easily translated into topics of discussion or narration. Power, its actors, its victims, and its mechanisms are often the talk of the town, and our everyday conversations, our mass media, our creative arts gladly use power as themes or motifs in discourses on society at large. Few stories are juicier than those of a president brutally abusing his power for his own personal benefit or for his own personal wrath against competitors for power - All the President's Men was a great movie. Few individuals are more fascinating than those who embody and emanate absolute power and are not afraid of wielding it in unscrupulous ways - Stalin, Napoleon, Mobutu, W. R. Hearst, and Onassis were all culture heroes of some sort in their days and afterwards. And scores of scholars ranging from Plato over Hobbes, Machiavelli, Marx, Gramsci to Foucault and Althusser have all theorised on the nature of power. Thus, we seem to have a strangely ambivalent attitude towards power: it attracts as well as repels; it fascinates and abhors at the same time; it has a beauty as well as an ugliness to it that match those of few other phenomena.
This book intends to offer a proposal for critical reflection on, and analysis of, discourse, and right from the start I wish to establish that a critical discourse analysis should not be a discourse analysis that reacts against power alone. It is a commonplace to equate 'critical approaches' with 'approaches that criticise power'. My point of view is that we need to be more specific. The suggestion I want to offer is that it should be an analysis of power effects, of the outcome of power, of what power does to people, groups, and societies, and of how this impact comes about. The deepest effect of power everywhere is inequality, as power differentiates and selects, includes and excludes. An analysis of such effects is also an analysis of the conditions for power - of what it takes to organise power regimes in societies. The focus will be on how language is an ingredient of power processes resulting in, and sustained by, forms of inequality, and how discourse can be or become a justifiable object of analysis, crucial to an understanding of wider aspects of power relations. I situate my argument in a particular environment: that of the present world system, that of so-called 'globalisation'. A critical analysis of discourse, I shall argue, necessarily needs to provide insights in the dynamics of societies-in-the-world.
In order to substantiate this, three central notions require clarification. The first one is the concept of discourse, our object of analysis; the second is the social nature of discourse; and the third is the object of critique in a critical analysis of discourse.
In this book, discourse will be treated as a general mode of semiosis, i.e. meaningful symbolic behaviour. Discourse is language-in-action, and investigating it requires attention both to language and to action (Hanks 1996). There is a long tradition of treating discourse in linguistic terms, either as a complex of linguistic forms larger than the single sentence (a 'text') or as 'language-in-use', i.e. linguistic structures actually used by people - 'real language' (Brown and Yule 1983; and de Beaugrande and Dressler 1981). This conception of discourse, broadly speaking, underlies the development of contemporary linguistic pragmatics. It has informed numerous studies in which, little by little, old and well-established concepts and viewpoints from linguistics were traded for more dynamic, flexible, and activity-centred concepts and viewpoints (Verschueren 1995, 1998; Verschueren et al. 1995; Mey 1998). This development was fuelled, on the one hand, by developments within linguistic theory itself, which called for more activity-centred approaches to analysis, the recognition of language-in-use as a legitimate object of analysis, and the discovery of grammatical and structural features of language operating at levels higher than the single sentence - coherence and cohesion (Halliday and Hasan 1976; Tannen 1984). On the other hand, it was fuelled by intensified interdisciplinary contacts between linguists and scholars working in fields such as literary analysis, semiotics, philosophy, anthropology, and sociology, where conceptions of language were used that derived from Boas, Sapir, Bakhtin, Saussure, and Jakobson (Hymes 1983). It was the (re)discovery of a radically different parallel stream of conceptions of language and analytical tools of analysing them that led to more mature approaches to discourse (Jaworski and Coupland 1999 provide a useful overview; see also Hanks 1989, 1996).
I intend to follow this pragmatic stream, but I also intend to widen it by including conceptions of discourse that could be called fully 'non-linguistic', in the sense that they would not be acceptable to most linguists as legitimate objects of inquiry. Discourse to me comprises all forms of meaningful semiotic human activity seen in connection with social, cultural, and historical patterns and developments of use. Discourse is one of the possible names we can give to it, and I follow Michel Foucault in doing so. What is traditionally understood by language is but one manifestation of it; all kinds of semiotic 'flagging' performed by means of objects, attributes, or activities can and should also be included for they usually constitute the 'action' part of language-in-action. What counts is the way in which such semiotic instruments are actually deployed and how they start to become meaningful against the wider background mentioned above. Recent semiotic work has shown how rather than single objects and instruments, intricate connections between all kinds of semiotic modes and media make up contemporary semiosis (Kress and van Leeuwen 1996). A typical newspaper advertisement nowadays contains written text in various shapes and formats, ranging from headlines to small print, with differences in shape or colour that are meaningful. It also contains images, pictures, logos, symbols, and so on; it is of a particular size and it displays a particular architecture - the overall makeup of such signs is visual rather than textual, or at least, the textual (content) cannot be separated from the visual (form). It occurs in a space-time frame: advertisements that are printed only once are different from those that appear every day over a period of time; those that appear on the front page have a different status from those that occur on page 6 of the paper. None of the components of the advertisement is arbitrary, but none of them is meaningful in itself: the object we call 'discourse' here is the total layout of the advertisement, the total set of features - in short, it is, not the text or the images. Contemporary discourse analysis has to account for such complex signs and needs to address them, first and foremost, as contextualised activities rather than as objects (Scollon 2001). So, though this book will offer primarily 'linguistic' materials, examples, and arguments, the wider set in which such items belong should not be lost out of sight. This is not a linguistic book.
The social nature of discourse
A second item that requires clarification is the social nature of discourse. Does discourse matter to people? Yes it does, and the clearest evidence for it is the simple fact that we use it all the time. It has been stated over and over again: the use of language and other meaningful symbols is probably what sets us apart from other species, and what accounts for the peculiar ways of living together we call society or community. There is no such thing as a 'non-social' use of discourse, just as there is no such thing as a 'non-cultural' or 'non-historical' use of it. But all of this is truistic; the full story is obviously far more complex and will require the remainder of this book to start being told. What concerns us here is how discourse can become a site of meaningful social differences, of conflict and struggle, and how this results in all kinds of social-structural effects. The fact is: it can, and does so all the time. The reason for this is that we have to use discourse to render meaningful every aspect of our social, cultural, political environment: an event becomes 'a problem' as soon as it is being recognised as such by people, and discursive work is crucial to this; a mountain becomes a 'beautiful' mountain as soon as someone singles it out, identifies it and comments on it to someone else. In short, discourse is what transforms our environment into a socially and culturally meaningful one. But this kind of meaning-construction does not develop in vacuo, it does so under rather strict conditions that are both linguistic (never call a mountain a 'bird' or a 'car') and sociocultural (there are criteria for calling something 'beautiful' or 'problematic'), and this set of conditions cannot be exploited by everyone in the same way. This is where social differences in discourse structure and usage emerge as a problem, something that invites investigation and precision. Again, this will make this book less 'linguistic' than social-scientific.
The object of critique
We need to specify what our object of critical investigation will be. My suggestion is that a critical analysis of discourse in contemporary societies is an analysis of voice. Voice is a complex concept with a considerable history of use in the works of, for example, Voloshinov (1973); Bakhtin (1981 1986); Ducrot (1996); and Hymes (1996) (see Thibault 1989; Roulet 1996), and with widely different definitions and modes of application. The way in which I shall use it in this book can be summarised as follows. Voice stands for the way in which people manage to make themselves understood or fail to do so. In doing so, they have to draw upon and deploy discursive means which they have at their disposal, and they have to use them in contexts that are specified as to conditions of use. Consequently, if these conditions are not met, people 'don't make sense' - they fail to make themselves understood - and the actual reasons for this are manifold. They will be the topic of the best part of this book. My point of departure is: in contemporary societies, issues of voice become ever more pressing, they become more and more of a problem to more and more people. Voice is the issue that defines linguistic inequality (hence, many other forms of inequality) in contemporary societies. An analysis of voice is an analysis of power effects - (not) being understood in terms of the set of sociocultural rules and norms specified - as well as of conditions for power - what it takes to make oneself understood. This will be my object of investigation; and needless to say this object is only partially linguistic in nature.
I am not saying anything new here; in fact, I align myself with a long and very respectable tradition in the study of language in society - we shall turn to this tradition below. I see my own contribution to this field as synthetic, as an attempt to bring together a number of insights and approaches that are dispersed over time, place, and subdisciplinary audiences. Bringing them together, however, may result in something new and perhaps more useful or more applicable. It is my firm belief that a wide variety of social-scientific disciplines could benefit from structured, disciplined attention to language and discourse (and, to be sure, I am not alone in this). But it is up to us, scholars of language, to do our jobs and to provide sound, tested, and practical tools for analysis to others (just as we may expect similar efforts from scholars in other disciplines). What follows is a modest attempt at providing such a tool.
1.2 THE CRITICAL POOL
Before moving on, I need to mark the space in which I shall situate myself. It is a space of ideas and scholarship that I find useful and relevant for this project: the critical pool from which I shall draw material and inspiration.
In recent years, Critical Discourse Analysis has become a household name in the social sciences, and the term - abbreviated as CDA - has come to identify a 'school' of scholarship led by people such as Norman Fairclough, Ruth Wodak, Teun van Dijk, Paul Chilton, and others. Largely grounded in a European tradition of scholarship, CDA has become a popular and firmly established programmatic approach to language in society with some institutional muscle. CDA was groundbreaking in establishing the legitimacy of a linguistically oriented discourse analysis firmly anchored in social reality and with a deep interest in actual problems and forms of inequality in societies. It also broke ground in its proclaimed attempt at integrating social theory in the analysis of discourse (see especially Fairclough 1992a; Chouliaraki and Fairclough 1999). And it produced a discourse about itself which was perceived by many as liberating, because it was upfront about its own, explicitly left-wing, political commitment. Consequently, many would now view CDA as synonymous with the critical study of language and discourse at large.1
Obviously, this is a mistake. CDA has done much to revitalise socially committed analysis of language after a long period in which the study of language was, and apparently had to be, a purely academic endeavour in the sense that problem-orientedness, let alone political agendas, were taboo if one were a linguist. And CDA has certainly done much to re-open the issue of how studies of language can, and should, be studies of society. Chapter 2 will expand on this. But CDA is one out of many attempts towards the development of critical approaches to language, culture, and society. In fact, it needs to be set against the background of a whole stream of such attempts throughout the twentieth century.
A comprehensive survey of such traditions would require a book of its own; it would also be burdened by terminological and ideological issues over what constitutes 'critical' and what does not. But to the extent that 'critical analysis' stands for performing analyses that would expose and critique existing wrongs in one's society - analyses that should be 'brought home' - there are quite a few candidates for that status. I would like to single two out because of their immediate relevance to the purpose of this book: American linguistic anthropology; and mainstream sociolinguistics. I am selecting these two not to create a contrast with CDA and even less as a suggestion of 'more and better' than CDA, but because it offers us two things. First, they will show us that CDA is part of a wider landscape of critical approaches to language and society, and will thus make our view of the contribution of CDA sharper and clearer. Second, they will offer us a number of theoretical principles of respectable age which we can use in the remainder of the book.
American linguistic anthropology
It is a commonplace to begin the story of American linguistic anthropology with Franz Boas, and, in fact, the move by Boas from the margin to the centre of American anthropology in the late nineteenth and early twentieth centuries marked the beginning of scientific anthropology as we now know it (Darnell 1998, 2001; Stocking, 1974; Hymes 1983). Central to the Boasian anthropological paradigm was cultural relativism, as we know. Boas and his students set out to investigate the 'Native point of view'; culture as seen, lived, and experienced by its members, and they underpinned this endeavour with epistemological and methodological arguments that deserve reiterating, even if they should by now be common knowledge.2 Two arguments in particular deserve our attention here.
First, Boas and his students saw the discovery (or, better, the (re)construction) of the 'Native point of view' as something that would provide, explicitly and implicitly, a critique of their own society. There was among the Boasians a widespread dissatisfaction with the way in which contemporary American society worked and lived. Providing descriptions and interpretations of alternative points of view articulated by Native American groups was sensed to contribute to the necessary revision of American mainstream culture. The superiority of this American culture was called into question by means of examples from cultural practices by groups whose culture was, in the climate of the time, defined as far inferior. Thus Edward Sapir (1924) would oppose the 'spurious' American culture witnessed in the 'efficient' but meaningless and unfulfilling routine practices of a phone operator to the 'genuine' culture of Native fishermen from the north-west coast, characterised by complex, meaningful, and culturally as well as individually satisfying practices. To Sapir (in a way remarkably applicable to present-day concerns), the uniformising tendencies of social values such as efficiency were devastating to 'genuine' culture (Darnell 2001: 119).
Second, the Boasians would emphatically abstain from passing value judgements on the cultural practices they observed, claiming that groups were fully operational, effective systems and that differences between groups were merely differences in 'standpoint' (Darnell 2001: 111ff.). Such differences represented different ways in which societies came to terms with their lives in a particular environment. This sense of completeness and efficacy, famously articulated in Boas' introduction to the Handbook of American Indian Languages (Boas 1911), extended to all aspects of a culture, from its religion to its linguistic system. Research into this internally coherent and homogeneous system involved a standpoint in its own right: anthropological research was biased by the position of the observer, and the Native point of view had to be distinguished from the anthropologist's point of view. Thus, thinking about other cultures and languages could no longer rely on presumed 'standards' or universal needs for all cultures, and '[a]nthropology offered its fellow social sciences a view from outside standpoints that otherwise were likely to persist without awareness of ethnocentrism' (Darnell 2001: 113).
What this amounted to was, in effect, a problemisation of difference as inequality. Ethnocentrism, as a standpoint deeply ingrained in scholarship and everyday thinking, was a denial of equivalence of standpoints that were functionally equivalent when observed in their particular environments. Anthropology emerged as a critically reflexive tool capable of exposing the dynamics of disqualification of alternative solutions to similar problems. Anthropology was as much about us as it was about Native American groups: the so-called Sapir-Whorf hypothesis, which claimed that groups saw, dissected, and acted upon reality very much in terms of the categories provided by their native languages, was not only about the Hopi but also about mainstream Americans, equally held in captivity by their own categories and ways of acting upon them.
What this amounted to, as well, was the foregrounding of contextual studies of cultural forms - what we would now call an ecology of cultural forms. An understanding of culture and language requires setting culture and language firmly in the whole of the system in which a group operates, and explaining culture and language not by reference to a universal standard but by reference to the particular environment in which this culture and language occurs. The principle of relativity entails contextualisation, a focus on concrete, actual ways of functioning of cultural forms.
Despite the gradual move from a holistic agenda towards more specialised forms of anthropology, there is a direct line in the American tradition of scholars emphasising these critical concerns, from Franz Boas, Edward Sapir, Ruth Benedict, Benjamin Lee Whorf, and Paul Radin over post-Second World War scholars such as Dell Hymes (e.g. Hymes 1996, 1969) and John Gumperz (e.g. Gumperz 1982) and later to anthropologists such as James Clifford (e.g. 1988), Johannes Fabian (e.g. Fabian 1983, 1986), Charles Briggs (e.g. Briggs 1996 1997; Bauman and Briggs 2003), James Collins (e.g. 1998), William Hanks (e.g. 1996), and many others. In the field of linguistic anthropology, this tradition has witnessed a growing concern for inequality and ideology in language, reflexivity in research, and the capacity of linguistic-anthropological research to address questions of immediate relevance to disenfranchised or vulnerable groups in society (see the collections by Brenneis and Macaulay 1986 and Duranti 2001; let it be noted that both John Gumperz and Dell Hymes actively contributed to this trend). Crucially important work has been done on the status of linguistic varieties, language variation, and language shift (Gal 1979; Hill and Hill 1980; Kulick 1992; Woolard 1989), on authority in language (see Bauman and Briggs 2003; the collections by Schieffelin, Woolard, and Kroskrity 1998; Kroskrity 2000; Gal and Woolard 2001; compare Milroy and Milroy 1985 and Cameron 1995), on narrative, literacy and schooling (Heath 1983; Collins and Blot 2003), on identity, discourse, and hegemony (Jaffe 1999), on discourse practices as constitutive of social identities (e.g. Conley and O'Barr 1990; Jacquemet 1996; Hall and Bucholtz 1995) and so on - concerns that sound familiar to those acquainted with CDA and indeed echo the programmatic concerns of CDA (e.g. Gumperz 1982; Woolard 1985; Irvine 1989; Gal 1989; Bauman and Briggs 1990). By anyone's standards, this tradition is critical, and I shall come back to it in various places in the next chapters.
There has not been much interaction between scholars from CDA and American linguistic anthropology, despite the fact that their programmes may very well be compatible and their agendas partially overlapping (Blommaert et al. 2001). Both traditions have nourished themselves on similar social-theoretical complexes (notably those developed by Foucault, Bourdieu, Bakhtin, and Voloshinov), as well as on similar technical-analytic paradigms such as conversation analysis or interactional sociolinguistics (compare e.g. Fairclough 1989 and Heller 1994). Yet, a few 'crossover' exceptions notwithstanding (e.g. Ron Scollon 1998, 2001), the general picture is one of two (or more) separate worlds - and a lot of untapped sources of mutual inspiration.3 There is far more critical work available than that which goes under the label of 'critical'.
Sociolinguistics has produced a remarkable body of such critical work and, in fact, one could argue that sociolinguistics arose out of a concern with differential distribution patterns of language varieties and forms of language use in societies - with difference and inequality in other words. There have been, and still are, various branches of sociolinguistics. One pole would be formed by a branch that has close affinities with the linguistic-anthropological tradition mentioned above (e.g. Gumperz and Hymes 1972; Bauman and Sherzer 1974; Hymes 1974a; Gumperz 1982) and focuses on interactional patterns in small communities and/or particular types of social encounters. The other pole would be a quantitative paradigm of variation studies, focused on the discovery of correlations between linguistic varieties and social variables such as race, class, or gender (e.g. Labov 1972; Sankoff 1980, 1988; Dittmar 1996). In between, there are branches that are strongly oriented towards sociology (Fishman 1972; Bernstein 1971) and branches that are very much linguistics-oriented (Halliday 1978), as well as several creative mixtures of various approaches (e.g. Eckert 2000). Closely related to sociolinguistics as a theoretical and descriptive paradigm are more applied branches such as, for example, studies of language planning (Fishman 1974) and bilingualism (Romaine, 1989; Hoffman 1991; Heller 1995) (see Meshtrie 2001 for a survey).
What ties these very diverse approaches together is a shared concern with the nature and distribution of linguistic resources in societies. And just like in the case of American anthropology, we can distil from sociolinguistics some general insights without which any critical endeavour in the field of language would be futile.
First, as for the nature of linguistic resources, sociolinguistics has demonstrated that 'languages' as commonly understood (i.e. things that have names such as 'English', 'French', 'Hindi', 'Zulu') are sociolinguistically not the most relevant objects. These 'languages' are, in actual fact, complex and layered collections of language varieties, and the study of language in society should not be, for instance, a study of English in society, but a study of all the different varieties that, when packed together, go under the label of 'English'. These varieties can be categorised on the basis of a set of parameters, including at least: (a) varieties identified on the basis of the modes or channels of communication: spoken versus written, direct versus indirect (mediated) communication, etc.; (b) geographically identified varieties - 'dialects', regional accents; (c) socially identified varieties often called 'sociolects' - class varieties, professional jargons, peer-group talk, age-, gender-, or ethnically marked varieties, etc.; (d) situationally or domain-identified varieties, i.e. varieties used on particular occasions or in particular social domains, such as peer-group talk, dinner table conversations, doctor-patient interactions, classroom interactions etc.; (e) styles, genres, formats of communication - formal versus informal varieties, storytelling, jokes, casual chat, public speech, media discourse, etc.
It is clear that every chunk of real language will carry all these features at the same time. As already said, there is no such thing as 'non-social' language: language manifests itself in society always and simultaneously in the shape of a package containing all of the diacritics mentioned above. Any utterance produced by people will be, for instance, an instance of oral speech, spoken with a particular accent, gendered and reflective of age and social position, tied to a particular situation or domain, and produced in a certain stylistically or generically identifiable format. And the point is: all of these diacritics are not only linguistic diacritics but also social ones. They reflect speakers' identities, expectations as to what speakers intend to accomplish in a particular act of communication, elements of the wider social structure in which speakers are caught, and so on.
© Cambridge University Press
Table of ContentsPreface; Acknowledgments; 1. Introduction; 2. Critical discourse analysis; 3. Text and context; 4. Language and inequality; 5. Choice and determination; 6. History and process; 7. Ideology; 8. Identity; 9. Conclusion: discourse and the social sciences; Notes; Appendix: English translations of the documents in chapter 5; Glossary; References; Index.