- Reprinted From Journal of Communication Spring 1983 Volume 33:2 Copyright \u00a9 1983 by The Annenberg School of Communications - Reprinted From Journal of...

This preview shows page 1 out of 23 pages.

You've reached the end of your free preview.

Want to read all 23 pages?

Unformatted text preview: Reprinted From Journal of Communication Spring 1983 Volume 33:2 Copyright © 1983 by The Annenberg School of Communications Discourse Analysis: Its Development and Application to the Structure of News by Teun A. van Dijk Concepts and contributions of the study of language to the explication of the structure and meaning of texts are reviewed and applied to the study of news. Developments in the last decade within such areas as text linguistics and, more generally , within the growing interdisciplinary study of discourse, have potential applications for the systematic analysis of mass media messages. Discourse analysis can make more explicit the classical approaches to "content analysis." It can also stimulate a research paradigm within mass communication that sees textual analysis not only as a method of research—for example, in the study of media effects—but also as an autonomous endeavor toward the construction of a sound theory of media discourse. Because the study of discourse has become a large field in the past ten years, my discussion here must be limited to those aspects of discourse analysis that seem most relevant for the study of media discourse. Thus, I will pay little attention to those properties of discourse that can be characterized in terms of linguistic grammar in the strict sense, such as the syntax and semantics of (isolated) sentences. Rather, I will be concerned with more specific textual structures that have been neglected in linguistics. Similarly, I cannot go into the details of stylistic and rhetorical analysis of media discourse, although much work in this área still needs to be done. Finally, I will also limit my application to news discourse in the press, thereby neglecting TV, film, and radio discourse, the role of images in audiovisual forms of discourse, and other types of newspaper discourse such as advertisements and commentaries. Teun A. van Dijk is Professor in the Department of General Literary Studies, Section of Discourse Studies, University of Amsterdam. 20 Discourse Analysis and the Structure of Nonos Of course, the study of discourse is not restricted to the structural analysis of texts. I will show that the psychology of discourse processing, concerned with the cognitive principies of the comprehension and remembering of texts, is of fundamental importance in mass communication research. The 1970s saw a wide interest in the study of "texts" or "discourses" in such disciplines as linguistics, semiotics, literary scholarship, psychology, sociology, and anthropology. In each of these disciplines, an interest in texts seemed to mark a paradigm shift with respect to earlier studies of the structures and functions of language. Besides a focus on the "system" of language, explicitly accounted for within structural and generative-transformational grammars (62), a new emphasis was placed on the analysis of language use (e.g., within cognitive and sociocultural contexts), on language use as social action (e.g., within the study of so-called speech acts [821), and on the analysis of "natural" data (e.g., everyday conversations). In several of these areas, attention shifted from the study of individual words, phrases, or sentences to an analysis of structures and functions of actual forms of language use, that is, to discourse. This wide interdisciplinary interest in the study of discourse has earlier historical antecedents, of course. Classical rhetoric, from the work of Aristotle to the present day, has always been concerned with the (persuasive) properties of discourse, and the sophistication of its analysis of rhetorical operations, such as the so-called "figures of style," has met with some competition only with the advent of structuralism (49, 58, 71). 21 Journal of Communication, Spring 1983 Structuralism also brought a decisive reorientation to the study of literary discourse (11, 47) in such areas as classical poetics and literary scholarship. This "structuralist revolution" in the classical study of discourse in the twentieth century has two main sources. First, structural linguistics provided the necessary methodological renewal by offering an explicit definition of structural units and categories and the formulation of rules. Second, anthropology, itself inspired by this development in linguistics, gave impetus to the very successful structural analysis of narrative (1, 4, 39, 72; for a survey, see 41). The systematic analysis of discourse would be unthinkable without these predecessors of the 1960s. Yet, in addition, a more general account of language use, interaction, and communication is needed. Despite its earlier emphasis on abstract sentence grammars, linguistics has been at the center of this development, providing the necessary explicit methods for the systematic analysis of discourse. Besides the mainstream paradigm of Chomskyan generative-transformational grammars, the 1960s saw the formation of several "schools" of linguistic discourse analysis (3, 12, 24). First, the tagmemic school, centered around the work of Pike (70) in describing non-Western languages, has always paid attention to discourse, especially to stories and the structure of paragraphs (40, 61). Second, among the European approaches to discourse in linguistics, the systemic grammar of Halliday has inspired many studies (10, 27, 44, 59) at the boundaries of grammar and stylistics and now manifests itself mainly in the school of discourse analysis at the University of Birmingham (8, 9, 83). The third major influence carne from those working in (mainly German) text linguistics and text grammar, who advocated the construction of grammars that would account for linguistic structures beyond the boundaries of the sentence (3, 12, 14, 67, 68, 69). The linguists also had company in other fields. Sociolinguistics urged that more attention be paid to actual language use and that the hitherto silently presupposed social nature of language be taken seriously. Apart from dialectal and sociolectal variations in language use, or the study of the interdependence of linguistic forms and social categories such as situation, institution, age, gender, status, and role or group membership, sociolinguistics also was confronted with language use in discourse, especially under the impetus of work by Labov (54, 55). It has been increasingly accepted that language systems and language use are not autonomous, but are inextricably related to the interactional functions and the social contexts of verbal communication: language and discourse forms thus mark or "indicate" their relevant social parameters (81) and are treated as manifestations of social action of a specific kind. Closely related to this development in sociolinguistics has been a growing interest in the study of many discourse forms (after the earlier study of myths and folktales) in their cultural context (2, 42) within the fields of anthropology and ethnography. Such studies typically show that 22 Discourse Analysis and the Structure of Netos storytelling not only has different structural categories in different cultures, but also puts specific constraints on who can tell what to whom under what circumstances or, similarly, on how greeting rituals or other speech events take place in such cultures. The emphasis on naturally occurring speech has led many to the analysis of everyday conversation. Inspired by earlier work in microsociology, mainly within the so-called "ethnomethodological" tradition (33), both linguists and sociologists have formulated a number of basic principles of the "dialogical" and interactional aspect of language use and discourse, such as turn-taking, strategic moves, and everyday storytelling and arguing (80, 85, 88). ,Ffoliwngarkyuecomprhnsi edbyth sentence-grammatical paradigm of generative-transformational grammar, psychologists, too, discovered discourse. Both cognitive psychologists and scholars from the new burgeoning discipline of Artificial Intelligence (AI) developed models for the processing of discourse and its representation in memory, which have been widely applied, for example, in educational psychology (e.g., in the study of reading). These psychologists not only formulated the important cognitive dimension of discourse interpretation, but they also developed their own structural models, such as narrative grammars, and were the first to work out explicitly the now well-known assumption that understanding discourse presupposes vast amounts of general knowledge of the world. Before bringing the theoretical and analytical results of these various approaches to discourse to bear on the study of masa communication, I will offer a brief summary of some basic principies for the analysis of discourse. Since the diversity of theoretical, methodological, and terminological persuasions is too impressive to allow a short synthesis, I will here formulate the main properties of discourse against the background of my own earlier work (12, 14, 16, 18). I refer to other work for details or different approaches. Verbal utterances, such as sentences, discourses, texts, or messages (I will henceforth use the term "discourse"), are usually analyzed first on different levels. The structures at each of these levels are accounted for by specific sub-theories or even sub-disciplines of linguistics. Thus, there is phonology, accounting for the structure of sounds and intonation, morphology, formulating the principles of word formation, syntax, providing the mles according to which words of different categories can be combined into grammatical sentences, and semantics, dealing with the meaning of words, phrases, sentences, or whole discourses by formulating tales of interpretation. As opposed to such "underlying" meaning structures, the phonological, morphological, and syntactic 23 Manual of Communication, Spring 1983 expressions manifesting this meaning are sometimes simply called "surface structures." In practice, much of the work in discourse analysis has concentrated on semantic structures, that is, on meaning, because earlier work on sentence grammars tended to focus on surface structures. In addition to there levels, different units of analysis can be distinguished in discourse: individual words (lexical items), various structures of the clause, whole sentences, sequences of sentences (paragraphs), or whole discourses. The overall topic or theme of a discourse, for instance, can be studied only at the semantic level of the discourse as a whole, not at the level of individual words or sentences. Hence, a rather rough distindtion is usually made between "local" and "global" structures of discourse, with the former pertaining to sentences and immediate sentence connections and the latter to larger segments of the discourse or the discourse as a whole. Next, cutting through the various levels mentioned aboye, and both locally and globally, different dimensions of analysis can be distinguished. Thus, stylistic variation can occur at several levels, such as lexical choice, word formation, or syntactic structures. Similarly, rhetorical operations (such as alliteration, parallelism, metaphor, or irony) also require definition on various levels. Finally, there are different modes of the manifestation and use of discourse, such as spoken or written/printed discourse, monologues, and dialogues. The various units, categories, dimensions, and levels, along with the rules defining them, will all be called "textual." However, discourses are not just isolated linguistic "objects," but are integral parts of communicative acts in some sociocultural situation, which I will call "context." Thus, it is a contextual property of the discourse type "verdict" or "plea" that it is rightfully used only in the courtroom and by a judge or lawyer. At the boundary of text and context, the pragmatic analysis of discourse is concerned with the dimension of action in which a discourse is taken as some conventional forro of social action (promise, threat, question, congratulation), called a "speech act." I have provided, in extremely succinct terms, some elementary notions of discourse analysis. The various schools of discourse analysis mentioned aboye can be distinguished, in part, on the basis of their specific interest in some textual or contextual property. Thus, some people will exclusively study discourse style, or intonation in spoken discourse, or overall meaning, or specific social constraints on the context. Similarly, there can also be specialization in certain discourse types or genres, such as everyday conversation, stories, classroom discourse, textbooks, proverbs, or news. Each discourse type, then, could—or rather, should—be characterized in terms of a specific combination of various textual and contextual properties. A judge's verdict, for instance, should have a specific (formal) style and is constrained to specific overall meanings (themes, topics). 24 Discourse Analysis and the Structure off News Thus, to focus on news discourse (as I will below) requires a full analysis of its various levels, units, dimensions, modes, and social contexts. Of course, a relevant analysis would focus on those structural aspects that are typical. Structural differences exist not only between two such disparate discourse types as an everyday conversation and a psychology textbook, but also between a spontaneous everyday talk and a job interview, although both are dialogues. It is therefore not easy to specify in general what the properties of a discourse are at the various levels and for the respective units and dimensions. Nevertheless, we can specify some fairly general characteristics, which then can be further detailed for news discourse. 1. Functionality. If a discourse is taken to be the utterance of a sequence of sentences in some social context, then the various properties of such a discourse are assumed to be functional with respect to various aspects of the social context. That is, both surface structures and meanings are produced and understood as indications about characteristics of the speaker (e.g., intentions, wishes, moods), the relations between speaker and hearer (e.g., confidence, intimacy, power), and the type of social situation (e.g., a court trial, a school lesson, a birthday party). This will hold for surface structure style, such as lexical choices and sentence structures, and also for the possible topics or themes talked about or the speech acts that may or should be performed with the utterance of the discourse. The functionality also holds, therefore, "within" the discourse: the surface structure not only expresses or indicates social structure, but also, and even primarily, is meant to express underlying meaning (35). 2. Meaningfulness. A textual sequence of sentences is distinct from an arbitrary collection of sentences in the sense that, in principle, such a sequence should be meaningful. One of the typical conditions for meaningfulness of a discourse is some kind of unity, which is usually described in terms of local or global coherence. Local coherence means that subsequent clauses and sentences are meaningfully related, because the facts to which they refer are causally related or because the propositions expressed by these clauses or sentences are related (one proposition may be a specification, generalization, or example of a previous proposition). Global coherence pertains to larger parts of the discourse; this kind of global unity is usually described in terms of such notions as "topic" or "theme." Such themes or topics are accounted for theoretically in terms of so-called "semantic macrostructures." Thus, a fragment of a discourse or a whole discourse is considered to be globally coherent if a topic (represented by a macroproposition) can be derived from such a fragment. Note that part of the meaningfulness criterion for discourse is not only that (sequences of) sentences have meaning, but also that they are "about" something: they refer to (real or imagined) 25 Discourse Analysis and the Structure of News course is less studied as a form of "social practice" in its own right, for which it is a legitimate aim to make explicit the inherent structures at all levels of analysis. Second, a discourse analysis primarily aims at the explication of qualitative data rather than quantitative data. Of course, quantitative measures may well be based on an explicit analysis of a more qualitative kind. Third, while content analysis is primarily based on observable, countable data, such as words, phrases, sentences, or stylistic features, a discourse analysis will—apart from making explicit such surface structures in tercos of modem grammars—also pay attention to underlying semantic structures and make explicit implications, presuppositions, connections, strategies, etc., which usually remain implicit in the discourse. It will try, in terms of empirical theories, to find the rules or principles underlying the structures, the production, and the comprehension of media messages. Finally, a discourse analysis will be part of a more embracing cognitive and social theory about the rules and strategies that underlie the production and understanding of media discourse. Instead of merely correlating, it will try to explain, in precise cognitive models, how various structures of media discourse come about and how media discourse is understood and represented in memory. Hence, the relation between content properties and specific "effects" is split up in terms of a number of highly complex cognitive and social-psychological models of information processing. Of course, these distinctive features of discourse analysis are taken to be relative to content analysis as a whole. There are certainly studies that have come close to one or more of these aims. Many of these studies have appeared in England (and in Germany; see 84) and often explicitly mention their opposition to traditional communications research in the U.S. Work done at the Centre for Contemporary Cultural Studies of the University of Birmingham (e.g., 43), influenced by several theorists in France (such as Barthes, Althusser, Poulantzas, Lacan, and Pécheux), has especially focused on the ideology of media discourse. Through more explicit linguistic discourse analysis (and another theory about the role of the media in society), such work attempts to uncover implied meanings that represent ideological positions. A similar goal can be found in the media analyses of some linguists (e.g., 28, 53) who, by meticulous syntactic analysis, show that the very sentence structures in news can mask who are the responsible "agents" in some events and that newspapers (e.g., the Sun and the Morning Star) thus can be differentiated linguistically according to their underlying ideologies. Finally, the Glasgow University Media Group (36, 37), in its well-known Bad News studies of television news, also applies a systematic verbal and visual analysis, uncovering how the very photographs or noun phrases employed can bias the news, for example, in favor of employers and against 27 Journal of Communication, Spring 1983 (striking) workers. These few studies indicate the kind of approach I hope to stimulate and, if possible, to make even more systematic and explicit (for detail, see 20). The theses formulated aboye can be illustrated and further elaborated for a specific kind of media discourse—news in the daily press. I will focus here on textual structures of news and largely ignore the various contextual conditions and constraints on such textual structures. The latter have received much attention in recent years, especially in sociology (7, 26, 32, 38, 76, 87). These studies emphasize that news is not simply an (incomplete) description of the facts, but a specific kind of (re)construction of reality according to the norms and values of some society. They also show that news production is part of a complex of professional routines f...
View Full Document

  • Fall '17
  • jane smith
  • Poetics, Speak, Journal of communication

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern

Stuck? We have tutors online 24/7 who can help you get unstuck.
A+ icon
Ask Expert Tutors You can ask You can ask You can ask (will expire )
Answers in as fast as 15 minutes