An Introduction to Critical Discourse Analysis in Education

Rebecca Rogers

• Publisher: Routledge
• Year: 2011
• Print ISBN: 9780415874298
• eBook ISBN: 9781136861475
• Topic: Education
• Subtopic: Education General

Accessible yet theoretically rich, this landmark text introduces key concepts and issues in critical discourse analysis and situates these within the field of educational research. The book invites readers to consider the theories and methods of three major traditions in critical discourse studies – discourse analysis, critical discourse analysis, and multimodal discourse analysis -- through the empirical work of leading scholars in the field. Beyond providing a useful overview, it contextualizes CDA in a wide range of learning environments and identifies how CDA can shed new insights on learning and social change. Detailed analytic procedures are included – to demystify the process of conducting CDA, to invite conversations about issues of trustworthiness of interpretations and their value to educational contexts, and to encourage researchers to build on the scholarship in critical discourse studies.

This edition features a new structure; a touchstone chapter in each section by a recognized expert (Gee, Fairclough, Kress); and a stronger international focus on both theories and methods.

NEW! Companion Website with Chapter Extensions; Interviews; Bibliographies; and Resources for Teaching Critical Discourse Analysis.

Part I Discourse Analysis

DOI: 10.4324/9780203836149-2

2 Discourse Analysis: What Makes It Critical?

James Paul Gee

Arizona State University

DOI: 10.4324/9780203836149-3

Introduction

This paper is about my approach to discourse analysis (Gee, 2005, 2007). It is also about the question: What makes discourse analysis “ critical discourse analysis”?

At the outset, I want to make a distinction that is important from a linguistic point of view: a distinction between utterance-type meaning and utterance-token meaning (Levinson, 2000). Any word, phrase, or structure has a general range of possible meanings, what we might call its meaning range. This is its utterance-type meaning. For example, the word “cat” has to do, broadly, with the felines, and the (syntactic) structure “subject of a sentence” has to do, broadly, with naming a “topic” in the sense of “what is being talked about.”

However, words and phrases take on much more specific meanings in actual contexts of use. These are utterance-token meanings or what I will call “situated meaning.” Thus, in a situation where we say something like “The world’s big cats are all endangered,” “cat” means things like lions and tigers; in a situation where we are discussing mythology and say something like “The cat was a sacred symbol to the ancient Egyptians,” “cat” means real and pictured cats as symbols; and in a situation where we are discussing breakable decorative objects on our mantel and say something like “The cat broke,” “cat” means a statue of cat.

Subjects of sentences are always “topic-like” (this is their utterance-type meaning); in different situations of use, subjects take on a range of more specific meanings. In a debate, if I say “The constitution only protects the rich,” the subject of the sentence (“the constitution”) is an entity about which a claim is being made; if a friend of yours has just arrived and I usher her in saying “Mary’s here,” the subject of the sentence (“Mary”) is a center of interest or attention; and in a situation where I am commiserating with a friend and say something like “You really got cheated by that guy,” the subject of the sentence (“you”) is a center of empathy (signaled also by the fact that the normal subject of the active version of the sentence—“That guy really cheated you”—has been “demoted” from subject position through use of the “get-passive”).

The Utterance-Type Meaning Task

Discourse analysis of any type, whether critical or not, can undertake one or both of two tasks, one related to utterance-type (general) meaning and one related to situated meaning. One task, then, is what we can call the utterance-type meaning task. This task involves the study of correlations between form and function in language at the level of utterance-type meanings (general meanings). “Form” here means things like morphemes, words, phrases, or other syntactic structures (e.g., the subject position of a sentence). “Function” means meaning or the communicative purpose a form carries out.

The other task is what we can call the utterance-token meaning or situated meaning task. This task involves the study of correlations between form and function in language at the level of utterance-token meanings. Essentially, this task involves discovering the situation-specific or situated meanings of forms used in specific contexts of use.

Failing to distinguish between these two tasks can be dangerous, since very different issues of validity for discourse analysis come up with each of these tasks, as we will see below. Let me start with an example of the utterance-type meaning task. Specific forms in a language are prototypically used as tools to carry out certain communicative functions (that is, to express certain meanings). For example, consider the sentence labeled (1) below (adapted from Gagnon, 1987, p. 65):

1. Though the Whig and Tory parties were both narrowly confined to the privileged classes, they represented different factions and tendencies.

This sentence is made up of two clauses, an independent (or main) clause (“they represented different factions and tendencies”) and a dependent clause (“Though the Whig and Tory parties were both narrowly confined to the privileged classes”). These are statements about form. An independent clause has as one of its functions (at the utterance-type level) that it expresses an assertion; that is, it expresses a claim that the speaker/writer is making. A dependent clause has as one of its functions that it expresses information that is not asserted, but, rather, assumed or taken-for-granted. These are statements about function (meaning).

Normally (that is, technically speaking, in the “unmarked” case), in English, dependent clauses follow independent clauses. Thus, the sentence (1) above might more normally appear as: “The Whig and Tory parties represented different factions, though they were both narrowly confined to the privileged classes.” In (1) the dependent clause has been fronted (placed in front of the whole sentence). This is a statement about form. Such fronting has as one of its functions that the information in the clause is thematized (Halliday, 1994), that is, the information is treated as a launching-off point or thematically important context from which to consider the claim in the following dependent clause. This is a statement about function.

To sum up, in respect to form-functioning mapping at the utterance-type level, we can say that sentence (1) renders its dependent clause (“Though the Whig and Tory parties were both narrowly confined to the privileged classes”) a taken-for-granted, assumed, unargued for (i.e., unasserted), though important (thematized) context from which to consider the main claim in the independent clause (“they represented different factions and tendencies”). The dependent clause is, we might say, a concession. Other historians might prefer to make this concession the main asserted point and, thus, would use a different grammar, perhaps saying something like: “Though they represented different factions and tendencies, the Whig and Tory parties were both narrowly confined to the privileged classes.”

At a fundamental level, all types of discourse analysis involve claims (however tacitly they may be acknowledged) about form-function matching at the utterance-type level. This is so because, if one is making claims about a piece of language, even at a much more situated and contextualized level (which we will see in a moment), but these claims violate what we know about how form and function are related to each other in language at the utterance-type level, then these claims are quite suspect, unless there is evidence that the speaker or writer is trying to violate these sorts of basic grammatical relationships in the language (e.g., in poetry).

As I have already said, the meanings with which forms are correlated at the utterance-type level are rather general (meanings like “assertion,” “takenfor-granted information,” “contrast,” etc.). In reality, they represent only the meaning potential or meaning range of a form or structure, as we have said. The more specific or situated meanings that a form carries in a given context of use must be figured out by an engagement with our next task, the utterance-token or situated meaning task.

The Situated Meaning Task

A second task that any form of discourse analysis, critical or otherwise, can undertake is what I called above the utterance-token or situated meaning task. For simplicity’s sake, I will now just call this “the situated meaning task.” When we actually utter or write a sentence it has a situated meaning (Gee, 2004, 2005). Situated meanings arise because particular language forms take on specific or situated meanings in specific different contexts of use.

Consider the word “coffee” as a very simple example of how situated meaning differs from utterance-type meaning. “Coffee” is an arbitrary form (other languages use different sounding words for coffee) that correlates with meanings having to do with the substance coffee (this is its meaning potential). At a more specific level, however, we have to use context to determine what the word means in any situated way. In one context, “coffee” may mean a brown liquid (“The coffee spilled, go get a mop”); in another one it may mean grains of a certain sort (“The coffee spilled, go get a broom”); in another it may mean containers (“The coffee spilled, stack it again”); and it can mean other things in other contexts, such as berries of a certain sort, a certain flavor, or a skin color. We can even use the word with a novel situated meaning, as in “You give me a coffee high” or “Big Coffee is as bad as Big Oil as corporate actors.”

To see a further example of situated meanings at work, consider sentence (1) again (“Though the Whig and Tory parties were both narrowly confined to the privileged classes, they represented different factions”). We said above that an independent clause represents an assertion (a claim that something is true). But this general form-function correlation can mean different specific things in actual contexts of use, and can, indeed, even be mitigated or undercut altogether.

For example, in one context, say between two like-minded historians, the claim that the Whig and Tory parties represented different factions may just be taken as a reminder of a “fact” they both agree on. On the other hand, between two quite diverse historians, the same claim may be taken as a challenge (despite YOUR claim that shared class interests mean no real difference in political parties, the Whig and Tory parties in 17th-century England were really different). And, of course, on stage as part of a drama, the claim about the Whig and Tory parties is not even a “real” assertion, but a “pretend” one.

Furthermore, the words “privileged,” “contending,” and “factions” will take on different specific meanings in different contexts. For example, in one context, “privileged” might mean “rich,” while in another context it might mean “educated” or “cultured” or “politically connected” or “born into a family with high status” or some combination of the above or something else altogether.

To analyze Gagnon’s sentence or his whole text, or any part of it, at the level of situated meanings—that is, in order to carry out the situated meaning task—would require a close study of some of the relevant contexts within which that text is placed and which it, in turn, helps to create. This might mean inspecting the parts of Gagnon’s text that precede or follow a part of the text we want to analyze. It might mean inspecting other texts related to Gagnon’s. It might mean studying debates among different types of historians and debates about educational standards and policy (since Gagnon’s text was meant to argue for a view about what history ought to be taught in schools). It might mean studying these debates historically across time and in terms of the actual situations Gagnon and his text were caught up in (e.g., debates about new school history standards in Massachusetts, a state where Gagnon once helped write a version of the standards). It might mean many other things, as well. Obviously, there is no space in a chapter of this scope to develop such an analysis.

The issue of validity for analyses of situated meaning is quite different than the issue of validity for analyses of utterance-type meanings. We saw above that the issue of validity for analyses of utterance-type meanings basically comes down to choosing and defending a particular grammatical theory of how form and function relate in language at the level of utterance-type meanings, as well as, of course, offering correct grammatical and semantic descriptions of one’s data. On the other hand, the issue of validity for analyses of situated meaning is much harder. In fact, it involves a very deep problem known as “the frame problem” (Gee, 2005).

The Frame Problem

The frame problem is this: Any aspect of context can affect the meaning of an (oral or written) utterance. Context, however, is indefinitely large, ranging from local matters like the positioning of bodies and eye gaze, through people’s beliefs, to historical, institutional, and cultural settings. No matter how much of the context we have considered in offering an interpretation of an utterance, there is always the possibility of considering other and additional aspects of the context, and these new considerations may change how we interpret the utterance. Where do we cut off consideration of context? How can we be sure any interpretation is “right,” if considering further aspects of the context might well change that interpretation?

Let me give an example of a case where changing how much of the context of an utterance we consider changes significantly the interpretation we give to that utterance. Take a claim like “Many children die in Africa before they are five years old because they get infectious diseases like malaria.” What is the appropriate amount of context within which to assess this claim? We could consider just medical facts, a narrow context. And in the context the claim seems unexceptional.

But widen the context and consider the wider context described below:

Malaria, an infectious disease, is one of the most severe public health problems worldwide. It is a leading cause of death and disease in many developing countries, where young children and pregnant women are the groups most affected. Worldwide, one death in three is from an infectious or communicable disease. However, almost all these deaths occur in the non-industrialized world. Health inequality effects not just how people live, but often dictates how and at what age they die. [see http://www.cdc.gov/malaria/impact/index.htm and http://ucatlas.ucsc.edu/cause.php]

This context would seem to say that so many children in Africa die early not because of infectious diseases but because of poverty and economic underdevelopment. While this widening of the context does not necessarily render the claim “Many children die in Africa before they are five years old because they get infectious diseases like malaria” false, it, at least, suggests that a narrow construal of “because” here (limiting it to physical and medical causes) effaces the workings of poverty and economics.

The frame problem is both a problem and a tool. It is a problem because our discourse analytic interpretations (just like people’s everyday interpretations of language) are always vulnerable to changing as we widen the context within which we interpret a piece of language. It is a tool because we can use it—widening the context—to see what informatio...

Table of contents

• Cover Page
• Half Title Page
• Title Page
• Copyright Page
• Dedication
• Table Of Contents
• Foreword
• Preface
• Critical Approaches to Discourse Analysis in Educational Research
• Part I Discourse Analysis
• Discourse Analysis: What Makes It Critical?
• Narratives of Exclusion and the Construction of the Self
• A Critical Discourse Analysis of Neocolonialism in Patricia McCormick's Sold
• Figured Worlds and Discourses of Masculinity: Being a Boy in a Literacy Classroom
• Part II Critical Discourse Analysis
• Semiotic Aspects of Social Transformation and Learning
• Learning as Social Interaction: Interdiscursivity in a Teacher and Researcher Study Group
• Language, Power, and Participation: Using Critical Discourse Analysis to Make Sense of Public Policy
• Locating the Role of the Critical Discourse Analyst
• Part III Multimodal Discourse Analysis
• Discourse Analysis and Education: A Multimodal Social Semiotic Approach
• Discourse in Activity and Activity as Discourse
• Mapping Modes in Children‘s Play and Design: An Action-oriented Approach to Critical Multimodal Analysis
• The Discourses of Education Management Organizations: A Political Design
• Biography
• Index