Jump to ContentJump to Main Navigation
Reframing Public PolicyDiscursive Politics and Deliberative Practices$

Frank Fischer

Print publication date: 2003

Print ISBN-13: 9780199242641

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/019924264X.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 18 December 2018

Public Policy and Discourse Analysis

Public Policy and Discourse Analysis

(p.73) 4 Public Policy and Discourse Analysis
Reframing Public Policy

Frank Fischer (Contributor Webpage)

Oxford University Press

Abstract and Keywords

This is the first of two chapters that present a more detailed examination of the ways in which social constructions are produced and negotiated in public politics through the medium of discourse. It looks at public policy and discourse analysis, and starts by defining discourse. The remaining sections of the chapter are: Discourse Analysis; Policy Discourse and Argumentative Struggle; Policy Storylines, which are the basic linguistic mechanism for creating and maintaining discursive order –– a generative sort of narrative that allows actors to draw upon various discursive categories to give meaning to specific or social phenomena (for example, ‘there is nothing we can do’ or ‘we must take immediate action’); and an appendix, Discourse and Social Change: Commodifying Educational Policy.

Keywords:   discourse analysis, discourse, discursive politics, educational policy, policy discourse, policy storylines, public policy, social change, social constructions

Discourse is not easy to define. Many authors in the primary anti‐positivist traditions of discourse analysis—hermeneutics, post‐structuralism, and post‐Marxist—use the term differently, making it difficult to offer a fixed or commonly accepted definition. Various characteristics of discourse, however, are relatively clear. Discourse, as Howarth (2002: 9) writes, refers ‘to historically specific systems of meaning which form the identities of subjects and objects’. Discourse theory, as such, starts from the assumption that all actions, objects, and practices are socially meaningful and that these meanings are shaped by the social and political struggles in specific historical periods. Through a range of linguistic and non‐linguistic materials—verbal statements, historical events, interviews, ideas, politics, among others—the goal of discourse analysis is thus to show how these actions and objects come to be socially constructed and what they mean for social organization and interaction. Included among the methods for doing this are the tools of rhetorical analysis, hermeneutics, deconstruction, and genealogical approaches to discourse analysis. The discussion that follows draws on these various approaches without adhering to a particular school of analysis.

Clearly, then, discourse is more than just synonymous with discussion or talking. The meanings of the words used and the statements employed in a discourse depend on the social context in which they are uttered, including the positions or arguments against which they are advanced. At the level of everyday interaction, discourses represent specific systems of power and the social practices that produce and reproduce them. As MacDonell (1986: 2) illustrates the point by showing how discourses function in different social settings: the kinds of speech appropriate on the factory floor differ sharply from those in the corporate boardroom, just as doctors in a hospital speaks differently from their patients. Different social groups, especially groups with differing degrees of power and authority, may use the same words differently in their interpretations of social and political situations.

For Hajer (1995a: 44) discourse is ‘a specific ensemble of ideas, concepts, and categorizations that are produced, reproduced, and transformed to give meaning to physical and social relations’. As such, a discourse and discursive practices circumscribe the range of subjects and objects through which people experience the world, specify the views that can be legitimately accepted as knowledge, and constitute the actors taken to be the agents of knowledge. Through such discursive delimitations, as Shapiro (1981: 130) puts it, a discourse establishes ‘norms for developing conceptualizations that are used to understand the phenomenon’. Discursive practices, which Foucault took as his basic unit of analysis, are the widely held and oft‐repeated interpretations of social conduct that produce and affirm behaviours. Over time these (p.74) interpretations become unreflectively taken for granted; they are scarcely noted by the actors who employ them. As generally accepted presuppositions, they become embedded in the institutional deliberations and practices that produce and govern basic societal relations.

Focused on virtually all research concerned with language in its social and cognitive context, including scientific disciplinary discourses, legal discourse, everyday narrative discourse, and so on, discourse analysis has given rise to numerous approaches. Basic to these orientations, though, is a concern with defining the units of analysis in various modes of deliberation and inquiry (Meinhof 1993). A discourse, in this respect, is not just any collection of words or sentences. Rather it is an integration of sentences—spoken or written—that produces a meaning larger than that contained in the sentences examined independently. Each kind of discourse links the sentences that compose it according to distinct patterns of reasoning. Physics, for example, is one kind of discourse; social science another; and law still another. And there are, of course, many others, including everyday discourses.

Fairclough (1992) offers a method that combines elements of hermeneutic, post‐structural, post‐Marxist and rhetorical approaches to discourse analysis. Following his approach, a ‘discursive event’ can be at once be seen as text, discursive practice, or social practice (Fairclough 1992: 4). ‘The “text” dimension attends to language analysis of texts.’ Focus on this aspect of discourse is the primary activity of linguistics (Johnstone 2001). Here a basic task is to understand how a discourse links together its utterances according to distinctive discursive forms—for example, referential, narrational, persuasive, expressive, or poetical (Kinneavy 1971). A discursive practice, of particular interest to literary scholars, focuses the processes of text production and interpretation. The analysis of discursive practices involves, among other things, an examination of which types of discourses are employed in particular contexts, especially which discourses are privileged in particular areas of policymaking and which are excluded. And as social practice, discourse refers to issues such as the institutional and organizational circumstance of the discursive event and how that shapes the nature of the discursive practice (Fairclough 1992: 4). Here one would focus on the kinds of assumptions that underlie practices of particular policy institutions—for example, the way an office of education might be governed by a particular educational philosophy, or the way in which a particular methodological orientation might be embedded in the rules and procedures of a political science association and its official journal. This concern with social practices is a primary interest of the social scientist.

Whether directly expressed in discussion and debate among speakers, or embedded in the practices of societal institutions, discourse communicates on two basic levels: the broad cultural level and the everyday level of communicative interaction. Although perhaps somewhat simplistically, these two levels can be usefully thought of in terms of micro everyday (first‐order) and macro socio‐cultural (second‐order) discourse, which is the primary concern of this chapter. At this socio‐cultural level, macro discourse transmits basic values and gives cohesion to shared beliefs, similar to what Schmidt (2002: 210) describes as ‘communicative’ discourse in Chapter 2. (p.75) Among other things, it supplies society with basic stories that serve as models of behaviour, both positive and negative.

Examples of discourse at the macro level are not difficult to find. Perhaps the best example in Western countries is Christianity. In Christian cultures, the most profound illustration is the story of Jesus of Nazareth, which provides insights for all members of society, even including people who are not especially religious. Such discourses convey to a society its basic socio‐cultural identity, that is, where it comes from, how it got there, and what its goals and values are. Or take the case of modern‐day Israel. It exists in significant part because of the Holocaust and the story is told and retold to remind Jews of who they are and why they must defend themselves. In Germany, of course, the same story conveys a very different message of shame and guilt.

At this cultural level, discourses function epistemically to regularize the thinking of a particular period, including the basic organizing principles of social action (such as the rules of feudalism or capitalism). Functioning as deep socio‐linguistic structures, discourses organize the actors' understandings of reality without them necessarily being aware of it. In the Foucaultian sense of the terminology, such epistemic discourses have formative or constitutive power that structures basic social definitions, meanings, and interactions in a socio‐cultural system. As large encompassing systems of meaning embedded in and transmitted by culture, macro discourses constitute the ‘residua’ of a society's or group's collective memory. They do so primarily in the form of stories that can be taken as the engrams basic to our modes of thinking and action.

Consider a couple of examples. Basic to French history and culture is idea of the French nation as carrier of civilization to other parts of the world, an idea grounded in the French Revolution and extended by the far‐flung missions of Napoleon. This contrasts with the English tradition that emphasizes the establishment of a parliament and the rights of Englishmen. Whereas French political discourse stresses the universal rights of man, English political discourse is grounded in the particular rights of Englishmen and a concern about parliamentary encroachments. Such larger discursive themes serve as points of orientation and departure in more concrete political and everyday discourses in these different political cultures, especially when it comes to dealing with each other. Under these broadly structuring discourses, political stories and narratives are told which not only reflect these general systems of meaning but also work them out in the concrete practices of the everyday world of social action. In the politics of the European Union, for example, these differences offer underlying reasons why the British are reluctant to turn over their sovereign powers to a continental governing body (Schmidt 2001). For the French, they constitute grounds for advancing the European Union, so long as the French government is one of its leading directors.

An even more deep‐seated example of such cultural residua is found in the way common everyday phases of a language can carry and sustain more fundamental norms which speakers no longer recognize as such. For example, consider the expression ‘alles in Ordnung’ in the German language, which is regularly repeated in the course of the most mundane events. No one thinks about it, but the phrase no doubt has roots in the emphasis on discipline and authority that runs through (p.76) Prussian history. (A famous story is told of the Prussian king who fired his generals because they failed to follow his orders to the letter, although their actions enabled them to win the Franco‐Prussian war.) Contemporary Germans today do not attach such deep meaning to the regular repetition of ‘alles in Ordnung’, which can be likened to Americans casually asking someone ‘how it's going?’A discourse analysis would nonetheless take note that the phrase is still there and resonates, even unintentionally or unconsciously, with part of German history. In this way, it carries forward important aspects of a culture. One would not want to draw immediate consequences from such an utterance, but would nonetheless want to ask about the way such an expression still tends to influence contemporary cultural orientations, including important social constructs such as efficiency or cleanliness.1 Albeit always in new ways, a culture constantly reproduces itself. The process of reproduction is one of the ways a people show or reveal themselves.

Discourse Analysis

Discourse analysis in politics begins with the recognition that discourses are distributed across institutions. In addition to the dominant discourses, competing discourses struggle to gain recognition and power. A key task for the analyst is to account for the viewpoints and positions from which actors speak and the institutions and processes that distribute and preserve what they say. As MacDonell (1986: 2–3) writes, a discourse ‘may be identified by the institutions to which it relates and by the position from which it comes and which it marks out for the speaker’. The speaker's position, however, does not stand alone. It can only be understood ‘as a standpoint taken up by the discourse through its relation to another discourse’, as when the owners speak to the workers.

The discursive constitution of society, as Fairclough (1992) makes clear in his very useful explication of the theory and methods of discourse analysis, does not stem from a free play of ideas in people's heads. Rather, it emerges from practices that are rooted in and oriented to basic social structures and ideological practices, particularly the tensions and conflicts to which they give rise. This connects discourse to practices that constitute—and naturalize, sustain, and change—the basic social and political significations of the world. Politically, such practices not only construct but also support and alter specific power relations between collective entities (classes, groups, communities, and so on) through which they are expressed. The ideological significations of these entities, as Fairclough makes clear, are generated within power relations as part of the exercise of and struggle over power. Discourse in politics is thus not only an activity in a power struggle, but also a stake in it as well.

The question of the relations among discourse, power, and ideology is the subject of complicated debates. Indeed, the term ‘discourse’ was in part introduced to deal (p.77) with problematic issues related to particular theories of ideology.2 Here, though, there is no widely accepted statement in the literature. Although discourses are not ideologies, they do intersect with ideologies, which supply the words of a discourse with different meanings. Built into the various meanings of discursive practices, ideologies—as interpretive constructions of reality—contribute to the production or transformation of power relations and political domination (Fairclough 1992: 86–91). Pecheux (1982) argues that discourses, which have their basis in language, nonetheless are always uttered in the social sphere, which is constructed ideologically. They always find their social meanings by reference to an ideological position (MacDonell 1986: 43–55). Even though they are not ideologies per se, discourses will generally have an intentional or unintentional relationship or position to one. Drawing on Althusser (1971) and Pecheux (1982), as MacDonell (1986: 102) writes, a discourse ‘is pinned down where it serves as a weapon in an ideological struggle’. It is in these struggles that a discourse finds its social meanings.

A first‐rate example of the relationship of discourse to ideology is found in the academic discourse of economics. Western industrial societies are organized as capitalist systems; thus it is difficult to render any statement completely outside the realm of capitalist ideology. Nonetheless academic economics presents itself as a value‐free scientific discourse independent of capitalist or socialist ideologies. It purports to identify and describe casual economic relations, often portrayed as ‘natural’, and offers an apolitical way to discourse about them. Despite the fact that modern‐day ‘neo‐liberal’ capitalist discourse can arguably be defined as the hegemonic discourse of our time, many contemporary introductory economics texts never—or almost never—mention the world ‘capitalism’, thus extracting the discourse from its historical and ideological context. A socialist, however, is correct to identify the capitalist understandings embedded in these discourses. The key words of such the discourse—‘management’, ‘workers’, ‘profit’, and ‘corporations’, and so on—are interpreted to have very different meanings by socialist ideology. Where a capitalist reads the economic text and unreflectively takes the word ‘profit’ for granted, the socialist reads the same word against a concern for exploitation.

Because ideologies reflect basic material and social relations in a society, they also supply people with different social identities. The various groups in society thus orient themselves ideologically in different ways to particular discourses. To stay with economics, the British Labour Party, or at least the left wing of the party, keeps its distance from the free‐market discourses of contemporary economics. In this respect, as Pecheux (1982) has argued, people construct their consciousness in terms of three basic orientations to the dominant discourses and their ideological orientations. Some freely consent to and identify with the social image held out to them by (p.78) the dominant ideology, for example, working class or upper class (they are the ‘good subjects’). Others refuse the social categories that are offered, adopting a ‘counter‐identification’. These ‘trouble‐makers’ include criminals, but also often liberal reformers. The third group, Pecheux argues, takes the orientation of ‘disidentification’. Whereas the counter‐identification position defines itself in oppositional terms (‘what you call the political crisis’), this third group sets out a new discourse. Although the identifications established by the dominant ideology can never be entirely evaded, this orientation seeks to transform and displace them. A traditional example is the Marxist theory of capitalism; a contemporary illustration is radical ecology. It here that one can understand the revolutionary Sartre's argument that the trouble with being radical reformer is that the opposition always sets your agenda. The alternative, he argued, is to be a revolutionary and seek to set the agenda yourself.

Gramsci's concept of hegemony provides a fruitful way of conceptualizing and exploring these interrelated ideological and political aspects of discourse, in particular the ways they both contribute to and are shaped by wider processes of social and political change (Laclau und Mouffe 1985; Fairclough 1992). Hegemony, in Gramsci's (1971) work, emphasizes domination across the economic, ideological, cultural, and political domains of society. It refers to the power over society by one of the economically defined classes in political alliance with other societal forces. For Gramsci, though, this does not mean a fixed or firm social equation. Hegemony, as he defines it, is an unstable equilibrium that always remains partial and temporary. For this reason, the focus is on the strategies of political leadership required to sustain the hegemonic balance of forces. Constructing and maintaining the alliances necessary to sustain the equilibrium, as discursive power, involves integrating rather than simply dominating subordinate classes to win their consent. Through a combination of ideological means and material concessions, hegemonic politics emphasizes the constant struggle around the points of greatest instability between classes or groups, in an effort to build, sustain, or block alliances. Hegemonic struggle, as Fairclough (1992: 92) explains it, ‘takes place on a broad front, which includes the institutions of civil society (education, trade unions, family), with possible unevenness between different levels and domains’.

Gramsci (1971: 324) offers a conception of political groups as structured by the diverse ideologies implicit in their discursive practices. Given the sometimes conflicting or contradictory components that make up such understandings of groups, Gramsci speaks of their ‘strangely composite’ character. Basic to these composites is the way in which common‐sense knowledge serves as both ‘a repository of diverse effects of past ideological struggles’ and a regular target for restructuring in social and political struggles. In common‐sense knowledge, ‘ideologies become naturalized, or automatized’(Fairclough 1992: 92). On this view, ideologies are understood to be fields of ‘conflicting, overlapping, or intersecting currents or formations’, or what Gramsci referred to as an ‘ideological complex’. This necessitates an examination of the processes that articulate and restructure ideological complexes. In more recent years, a newer group of ‘neo‐Gramscian’ scholars have sought to revitalize Gramsci's insights in the area of international relations and foreign policy (Gill 1993; Cox 1994; Sinclair 2000).

(p.79) Such an approach to political struggle offers a dialectical perspective on the relationship between discursive structures and events. Viewing discourses as the relatively unstable and often contradictory configurations that hold together the equilibria of political hegemony, ‘the articulation and rearticulating of orders of discourse is correspondingly one stake in hegemonic struggle’ (Fairclough 1992: 93). Furthermore, ‘discursive practice, the production, distribution, and consumption (including interpretation) of texts, is a facet of hegemonic struggle which contributes in varying degrees to the reproduction or transformation not only of the existing order of discourse, but also to the reproduction of existing social and power relations.’

To illustrate the relationship between discourse and social change Fairclough offers a number of examples. His concern is with broad tendencies in discursive change affecting the societal order of discourse, which are related to more general directions of social and cultural change. One of his examples to explain the ways that language and discourse are used to facilitate social change is that of ‘commodification’. This involves analysing a process whereby social domains and institutions, whose concern is not producing economic commodities in the narrower economic sense of goods for sale, come nevertheless to be organized and conceptualized in terms of commodity production, distribution, and consumption. ‘In terms of orders of discourse’, he writes (1992: 207), ‘we can conceive of commodification as the colonization of institutional orders of discourse, and more broadly of the societal order of discourse, by discourse types associated with commodity production.’ Specifically, the process of commodification is used to show the way in which a language text can be employed to mediate a social transformation to a more entrepreneurial market‐oriented culture, a phenomenon of no minor importance. Since the Thatcher years in Britain and the Reagan period in the USA, such a process of commodification has been increasingly extended to one institutional domain after another.

Education in both countries is an important example. Towards this end, Fairclough concretely illustrates the role of the commodification discourse in mediating the process of social change through the example of the transformation of the traditional liberal educational curriculum to a more practical training orientation for the labour market. In the process, he shows how education is increasingly reconceptualized as an ‘industry’ or ‘enterprise’ concerned with producing and selling cultural or educational commodities to their student ‘clients’ or ‘consumers’. A widespread feature of contemporary educational discourse is the wording of courses or programmes of study as ‘commodities’ to be ‘packaged’ by curriculum designers and ‘marketed’ in a competitive bid for ‘customers’. (See Appendix.)

In order for a specific discourse to become hegemonic, according to Laclau and Mouffe (1985), a particular set of ‘nodal points’ must be established in the discursive field. Given the ‘surplus of meanings’ that characterize the discursive fields of most policy issues, as illustrated in Chapter 2, a candidate for hegemony must compete with a wide range of potentially relevant but different discourses. Within the discursive field, as Battistelli and Ricotta (2001: 4) write, ‘to become hegemonic the specific discourse needs to dominate the field in which it was formed, blocking the flow of differences in meaning and setting itself up as the center of (p.80) interpretive processes’. This happens as the discourse succeeds in the ‘structuring of privileged discursive points, or nodal points of discourse’ that facilitate discussion about some things and not others: that is, when its concepts and definitions have succeeded in fixing specific meanings to be employed in interpreting a wide spectrum of social issues, problems, objects, actors, and situations.

For Fraser (1989: 164) such discourse politics is understood in terms of struggles over the interpretation of issues rather than over substantive questions. This ‘politics of conflict interpretation, includes not only the interpretations of the problems, but also the types of discourses they should involve’ (Braun and Herrmann 2001: 2). This involves examining questions such as whose interpretation becomes authoritative, as well the reason why. Are the dominant forms of political discourse available for interpretive argumentation shown to be acceptable and fair to the parties involved? Do the arguments favour the dominant political groups or do they assist challengers in making their oppositional claims? To get at these questions the interpretive analyst has to examine the social and institutional logics that underlie the interpretive processes. In which societal institutions are the interpretations developed, and what forms of social relations bind interlocutors together?

Towards this end, discourse analysis must also include a focus on the available discursive resources in a given social situation and how they are stratified in accordance with societal patterns of dominance. Discursive resources involve accepted idioms in which claims can be advanced, such as talk about rights, or economic talk; available vocabularies, such as medical, religious, or feminist vocabularies, frameworks of argumentation (especially as they pertain to debates among citizens, experts, politicians, and others), and narrative approaches or conventions that are available for the construction of stories and their telling. Especially important in Fraser's scheme is an emphasis on modes of subjugation. How do different discourses position or subjugate participants in a communicative exchange (for example, are they experts or citizens, politicians or businessmen, and so on?).

Such an orientation would also require us to examine the distribution of discursive capabilities across a society (Beck 1986; Schiller and Symthe 1992). Some people can speak regularly, while others can hardly ever be heard, unless they threaten violence. A key aspect in modern society is the role of the media, particularly how it is controlled. As is well documented, the media in the United States and Europe is increasingly controlled by smaller numbers of people and the content is more and more influenced by the advertisers who pay for the programmes. This strongly supports the ideological complex with a distinctive consumer‐oriented business bias. Moreover, much of the news is limited in the perspectives it offers. Consequently, such research shows that modern citizens are socialized more into roles as passive members of the audience than active participants. From this perspective, education and access to information and information technologies, such as computers and the Internet, as resources becomes an important topic. Some writers such as Beck (1995) have, in this respect, described the defining tensions of the twenty‐first century in terms of a conflict between those who have expertise and those who do not.

(p.81) Fraser distinguishes between the dominant means communication and interpretation that are officially approved or sanctioned, and the subordinate discourses that are normally denied easy access to the primary discursive arenas. Discursive politics, from this perspective, is an essential strategy of political resistance (Fraser 1989: 165). The arguments of the subordinate groups, argues Fraser, cannot be ‘translated into the dominant discourses without being redefined in ways that shed important meanings’. Towards this end, Fraser calls for the development of ‘social spaces’ where groups with unequal discursive (and non‐discursive) resources can deliberate about different ways of talking about social issues and problems.

Discursive understanding of power and political struggles, as Fraser's Habermasian focus also makes clear, holds open the possibility of moving beyond the traditional negative connotations associated with ideology and power. In much of social science power and ideology are understood negatively; they refer to domination, manipulation, and control. Missing from this view, as we noted in Chapter 2, is the more positive understanding of discourse. Communicative power, as Habermas and others have made clear, can create new consenses that open the way to alternative identities and courses of action. Moving beyond domination or the mobilization of resources, discursive power is productive power. It is, as Litfin (1994) puts, it, ‘capacity‐giving’, a topic to which we return in later chapters.

Policy Discourse and Argumentative Struggle

Can we apply the discourse model more directly to the policy research? Such work is increasingly emerging. In Governing Molecules, for example, Gottweis (1998; 2002), has examined the discursive construction of genetic engineering in Europe and the United States to show that the genetic controversy is a process inseparable from the social construction of political, economic, and scientific worlds. Calling for a ‘poststructural policy analysis’, Gottweis employs discourse‐analytic methods based on insights from Derrida and Foucault to demonstrate the need to examine how discourses and narrative stories that create orientation and meaning constitute the policy field of genetic engineering. Critical of the assumption that political science can simply assume the ‘existence’ of actors and structures in politics, as well as the narrow understanding of politics that has informed conventional political studies, he argues for ‘conceptualizing the process of policymaking as situated at the intersection between forces and institutions deemed “political” and those apparatuses that shape and manage individual conduct in relations to norms and objectives but are deemed “nonpolitical,” such as science or education.’ Rather than ‘assuming stable boundaries between sectors such as politics, the economy, and science’, he calls for the study of ‘the micropolitics of boundary drawing’ (Gottweis 1998: 27).

Through this perspective Gottweis (1998) shows that the politics of genetic engineering cannot be reduced to citizens demanding the regulation of recombinant DNA technology. The politics of genetic engineering also takes place when scientists carry out specific experiments in genetic engineering and subsequently make claims about the importance of their research for the understanding of both nature and (p.82) social behaviour. This approach allows Gottweis to identify a much more sophisticated regime of ‘governability’, a system of fields and sites ranging from research laboratories to legislatures that negotiate and deploy approaches designed to control and manipulate genetic materials, or as he puts it, ‘to make them governable’.

In Ozone Discourses, Litfin (1994: 3) offers a discourse or ‘reflective’ approach to the 1987 negotiation of the Montreal Protocol (on substances that deplete the ozone layer). Essentially, it is a story about how a dominant anti‐regulatory discourse succeeded in supplanting a new environmental regulatory discourse. Through the narrative discourse she illustrates how mainstream functionalist, agency‐centred, and epistemic community approaches downplay, even ignore, the ways in which the information of scientific experts mainly reinforces or reinforces existing political conflicts.

On the surface of the matter, this landmark ecological treaty would seem to be the outcome of a rigorous risk analysis process and astute diplomatic negotiations. With complex atmospheric models as the scientific basis for negotiation and decision‐making, this account has led writers such as Haas (1989; 1992) to explain the Montreal process in terms of an epistemic community composed of atmospheric scientists. In this view, knowledge is taken to be a body of concrete and objective facts that were not only relevant to the actors' perceived interests, but were fundamentally implicated in questions of interpretation and policy framing. Knowledge provided the means that reoriented actors' understandings of their own interests, thus making the agreement possible.

Litfin (1994: 46–51), in sharp contrast, offers a very different picture of what happened. The ‘knowledge’ at issue was itself ‘framed in light of specific interests and preexisting discourses so that questions of value were rendered as questions of fact, with exogenous factors shaping the political salience of various modes of interpreting that knowledge’, including the discovery of an ozone hole of the Antarctic that helped to empower the subordinate regulatory discourse. By neglecting issues of interpretation and legitimacy, mainstream theories offer explanations that fail to capture the multi‐dimensional interpretive and interactive processes that link science to politics. In so far as these reflective elements are essential to knowledge‐based power, she argues that standard accounts have to be extended to include a discursive, productive, or capacity‐giving conception of power.

In The Politics of Environmental Discourse, Hajer (1995a) has analysed the politics of the acid rain controversy to show how different applications and understandings of acid rain have generated very different types of politics in two countries. Because Hajer's concept of discourse coalitions most directly engages the analytical and methodological debates in policy analysis, particularly in terms of his critique of the influential work of Sabatier and Jenkins‐Smith on advocacy coalitions, we turn to a more detailed examination of his contribution in the section below.

Devoted to a comparison of acid rain policies in Britain and the Netherlands, Hajer offers a full‐scale study of discursive policy analysis. Building on Foucault's theoretical work, Hajer labours to translate this lofty theoretical contribution into a framework for the analysis of concrete policy problems. The task, as he explains, is to understand how particular discourses in a distinct political domain come to influence (p.83) discourses in other domains. Towards this end, he draws on discursive concepts developed by the social psychologist Billig designed for a less abstract ‘level of interpretive interaction’. The result is a ‘social‐interactive discourse theory’ geared to the analytical tasks of policy analysis.

In Hajer's scheme, as in Foucault's, the dominant discourses supply social actors with ‘subject‐positions’ that define their social and power relationships in terms of a principal narrative or narratives (1995a: 65). Discourse analysis shows the way in which actors attach meanings to other actors in terms of the positions assigned or attributed to them in the dominant or counter‐discourses. Actors are ‘positioned’ with regard to specific social attributes such as blame, and responsibility. In this view, social actors mainly make sense of the world by borrowing terms and concepts from the discourses made available to them in their social groups and the society generally. That is, to be understood and considered relevant, a speaker has to situate his remarks in—or relate them to—the recognized discourses in use at the time. A speaker's statements must say something in terms of the ongoing flow of discursive exchanges.

Ideas that do not draw on or interact with the available discourses will be dismissed as strange or irrelevant. Put more concretely, before Adam Smith came along and offered a fully developed discourse on the virtues of the free market, it was difficult, as economic historians show, for the small businessman to express or convey such interests and concerns. Now capitalist discourse has long been one of the available discursive themes to which one can appeal—approvingly or disapprovingly—in an effort to be understood. Today, one can say the same of environmental discourse. Forty years ago it was difficult to express concern about the environment without confronting furrowed brows. Now it is more than clear what is at stake when one refers his or her argument to environmental concerns.

Actors in this discourse‐analytic approach are constituted by discursive practices, meaning they have to be understood through the concepts and languages employed to describe their activities. In the case of the public administrator, for example, conventional social science typically defines a public administrator in terms of his or her functional role in the agency. But a discursive analysis would examine the other stories co‐workers tell that interpret the administrator's behaviour in different ways, for example as party functionary, dedicated man of the people, or small‐minded bureaucrat, or perhaps all three, depending on the occasion.

Rather than a set of ritualized roles, social interaction is thus conceptualized as an exchange of arguments, of competing, sometimes contradictory, suggestions of how one is to make sense of reality. For political and social research this means exploring discourse practices through which social actors seek to persuade others to see a particular situation or event. In this dialectical or perspectival view of knowledge, it is important to know the counter arguments and positions of the others involved (MacDonell 1986: 43–59). As discursive arguments always take their meaning in opposition to other positions, the analyst always has to pay attention to the dialectical tensions among rival positions, a point to which we return in Chapter 6. The discursive fields in which communicative exchanges take place thus need to be a central interest of analysis. ‘Language’, as Shapiro (1981: 14) puts it, ‘becomes part of data (p.84) analysis for inquiry, rather than simply a tool for speaking about an extra‐linguistic reality.’

Underlying this approach are two basic analytical tasks, one concerned with the substantive content of discursive statements and the other related to their implications for institutional practices. With regard to discursive content, one can easily grasp the point by comparing the discourses of physics and environmentalism. Physics is a discourse, produced, reproduced, and transformed through practices like academic teaching, laboratory experiments, and peer‐reviewed journals. Environmental discourse, in contrast, is produced through activities such as practising of alternative life‐styles, engaging in political demonstrations and protests, making references to pristine forests, reciting folklore about nature, and the rejecting consumer culture. Both discourses have their own context or domain. No one can successfully use the language or modes of argument of physics in everyday discourse. And the same holds true of the language of environmentalism; it does not get one very far in the physics laboratory.

Hajer's specific concern is with the discursive content of environmental policy discourse, yet another kind of discourse with its own characteristics. While we can identify a rational or systematic basis for policy discourse, it lacks the kind of coherence of content found in formal discourses such as physics or law. Legal discourse, for instance, is a highly structured, internally consistent mode of deliberation with its own inherent criteria of credibility. Its systematized institutionalized practices are reflected in the procedures of lawsuits, the preparation of legislation, appeals processes to higher courts, and the like. Policy discourse is also dependent on its institutional environment, but it cannot achieve the kind of internal coherence of a discourse such as law, or even political theory for that matter. This has to do with its position at the intersection of various discourses, combining a range of discursive components—empirical, institutional, pragmatic, and normative factors—that give it its unique character and requirements. In the case of regulatory policy, for instance, Jasanoff (1990; 1995) has captured this in her concept of ‘regulatory science’, which she juxtaposes to ‘science’ per se. Where the later follows the logic of empirical causality, regulatory science involves an interdisciplinary interaction between legal, economic, scientific, and administrative discourses. Each contributes to what can become consensually established as a set of rules designed to judiciously rule over a particular policy domain.

In Hajer's (1995a: 53) social‐interactive approach, the interplay of such discourses is not to be conceptualized as ‘semi‐static plays in which social actors have fixed and well memorized roles of such as environmentalist, policy‐maker, scientist, or industrialist’, or the firm belief systems stereotypically attached to these roles. To the contrary, politics becomes an argumentative struggle in which actors not only try to make their opponents see the problem according to their view, but also seek to position or portray other actors in specific ways (for example, by showing the relationship of a particular argument to the interests of an industrialist) that open up new possibilities for action. Grounded in qualitative methods, the discourse‐analytic task is geared to the real world of social action. The evidence for this is found in the (p.85) actors themselves. Discourse is not a category merely imposed on them by social scientists; rather, it is something social actors constantly practise.

A significant part of the task is to identify the elements in argumentative interactions that are essential to explaining the emergence and persistence of particular discursive constructions. How, for example, do concepts such as ‘sustainable development’, ‘acid rain’, or ‘globalization’ get constructed and shaped? What kinds of forces—both political and rhetorical—are brought to bear to establish such discursive formations, and how do they affect the behaviour of both individuals and social groups? Once established, how can such constructions establish discursive hegemony over a particular policy domain?

A social‐interactive discourse approach takes social actors to be actively engaged in choosing and adapting thoughts, shaping and fashioning them, in an ongoing struggle for argumentative triumph over rival positions. Taking here a position that would seem to draw as much or more from Habermas than Foucault, Hajer sees the task of the discursive policy analysis as uncovering the defining claims of a particular position. That is, we need to examine the structure, style, and historical context of an argument to determine why some modes of argumentation serve to effectively justify specific actions in particular situations and others fail. The discourse analyst needs to examine a particular conception of reality as something that is upheld by key political actors through discursive interaction, as well as how the oppositional forces seek to challenge these constructs.

Most fundamental to these discursive interactions is the question of how a particular framing of an issue can bestow the appearance of problematic on some features of a discussion while others seem proper and fixed. The task of the analyst is to show whether particular definitions ‘homogenize’ a problem, that is, render the problem understandable by situating it is a wider social frame, or whether definitions lead to a ‘heterogenization’ that opens up established discursive categories and hence the possibility of new courses of action. In the process, the opening up, or ‘deconstructing’, of a policy discourse can show how it emerges as the unintended consequence of a confluence of events and ideas. It can also show how seemingly technical issues can conceal normative commitments, as well as what sorts of institutional arrangements make this possible. Discourse analysis, in this way, helps us see which institutional dimensions are firmly entrenched and which structural elements are more open to change.

Important to understand is how social actors can contribute to the production and potential transformation of a discourse. In this respect, Hajer (1995a: 55) adopts an ‘immanentist’ view of language. Like Marx, Hajer argues that discursive actors can shape their own history, but not under conditions entirely of their own choosing. Given that they are operating in pre‐existing social structures, the political context of interaction has to be situated in a discursive construction that offers both possibilities and constraints. Oppositional groups must recognize that various modes of expression have meaning or effect only to the extent that they are appropriate to the context in which they are uttered. Because the rules and conventions that constitute the social order have to be constantly reproduced and reconfirmed in actual speech situations—whether in documents, media coverage, or public debates—the challenge for such (p.86) groups is to find ways to break open these routines. In general, this happens through the discovery of contradictions or paradoxes.

In view of the complex multi‐dimensionality of a given social configuration, the particular conceptions of reality that ruling groups advance to legitimize existing institutions must necessarily cover up contradictions and paradoxes. The task for the opposition is to open and exploit these ideological tensions and contradictions by showing how they function to hide or conceal other realities (Mannheim 1936). At times, such openings are facilitated by the emergence of exogenous crises or shocks to the social system that make these tensions either visible or difficult to conceal. However, unlike Sabatier's (1988) emphasis on external shocks, which tend to play a quasi‐deterministic role, the crucial issue from a discourse perspective is the kind of strategic communications to which the such a crisis gives rise.

Discursive responses to a crisis are in no way automatic. The ways in which they are portrayed can vary widely, with very different implications, ranging from ‘there is little that can be done; we must accept it’ to ‘dramatic action must be taken—those responsible for allowing it to happen must be punished’. Compare here the rhetoric of President Herbert Hoover during the Great Depression with that of President Franklin Delano Roosevelt. Where Hoover could only ring his hands in despair (‘There is nothing more we can do’), Roosevelt energized the nation to rally around a new set of institutional arrangements (‘There is nothing to fear but fear itself’). Without the strategic discursive responses of a dynamic new leader, as Houck (2001) shows, the course of American history would have surely been quite different. Nothing is given in such political developments. How discursive responses will unfold are case‐specific, depending on a wide range of factors from how those who have power choose to exercise it, to the personal characteristics of the particular actors on the stage at the time, what sort of orientation the media is disposed towards, the mood of the public, the resources of different groups, other related contextual factors, and more. The point is that a systems crisis itself does not determine what follows; rather, it is only the trigger for discursive responses that open or close off possible courses of action.

Policy Storylines

The basic linguistic mechanism for creating and maintaining discursive order, or responding to a destabilizing jolt to the discursive order, is the concept of the storyline. A storyline, Hajer (1995a: 56) writes, ‘is a generative sort of narrative that allows actors to draw upon various discursive categories to give meaning to specific or social phenomena.’(For example, ‘there is nothing we can do’ or ‘we must take immediate action’.) The primary ‘function of story‐lines is that they suggest unity in the bewildering variety of separate discursive components parts of a problem’ that otherwise have no clear or meaningful pattern of connections.

To understand the world around them, most people do not draw on comprehensive discursive systems for their cognitions. Instead, they rely on storylines. That is, they do not appeal to well‐developed theories of ecology or political philosophies for cognitive assistance, but rather turn to discursive storylines as short‐hand constructions. (p.87) Storylines, in this way, function to condense large amounts of factual information intermixed with the normative assumptions and value orientations that assign meaning to them. As social constructions of particular events, storylines serve to position social actors and institutional practices in ongoing, competing narratives. In the process, they (p.88) stress some aspects of an event and conceal or downplay others (covering over potential problems—tensions or paradoxes—that might be embedded in the story).

Maintaining the social order depends on the successful reproduction of particular understandings of the relevant actors and institutions—that is, the capitalist and the free market economy require legitimization and active participation. At the same time, critical tensions in the free‐market storyline (‘the capitalist is only interested in his profit margin’) coupled with the emergence of counter storylines based on new problematic events or revelations (such as a sudden and dramatic increase in oil prices) open the way for efforts to introduce stronger market regulations. Finding or reconstructing the appropriate storyline becomes a central form of agency for the political actor. Effective communicative skills and argumentative strategies are thus important resources that play a significant role in determining the shape of events. Like an art form, such communicative skills hold out the possibility of making people see things differently, and, in the process, shifting the course of political struggle.

The relevance of discursive strategies is particularly evident in environmental politics. The environmental movement has long been haunted by the dilemma of whether to argue on the terms set by the government or to insist on their own mode of expression. In the latter case, of course, the movement runs the risk of losing its direct influence and therefore often trades its expressive freedom for influence on the policymaking process.3 Given a hegemonic discourse, people who try to challenge the dominant storyline are often expected to position their contribution in terms of established categories. Indeed, this is the primary way a hegemonic discourse exercises it power. The concept of ‘sustainable development’ is a good example of the process.4 Because the earlier environmental storyline, ‘limits to growth’, proved to be a non‐starter for the industrial community (as capitalist industry exists to grow), there was a need to innovate a new storyline capable of working for both environmentalists and industrialists. The response was the concept of ‘sustainable development’. Shaped by leading industrials, it refers to environmental protection but without sacrificing industrial development, which environmentalists have long defined as the very source of pollution.

Environmentalists can—and often do—reject the terms set by the dominant discourse, insisting on the importance of their alternative discourse. The disciplinary (p.89) force of discursive practices, as Hajer (1995a) explains, often depends on the implicit assumption that subsequent speakers will respond within the same discursive frame. Such discursive challenges may consist of rejecting an understanding of events and actions in terms of established categories or, often even more powerfully, in establishing new combinations with the operative, routinized discursive structures. Indeed, this is what much of contemporary environmental politics is about.

Although this social‐interactive theory of discourse focuses on how people take up or deny particular positions, it also recognizes the ways in which power is embedded in the reifications of a discourse and, as such, remains hidden. People use particular terms without giving much thought to the way they sustain a specific social order. Merely consider the difference between saying ‘Negro’ and ‘Blackman’. Both words refer to the same physical person, but each rests on and reproduces a very different conception of power relations between the races. Many people simply assume that such a word is a formal name for such a person with a particular skin colour, or that it is ‘just the way one talks’ on particular occasions, without being explicitly aware of (or taking into account) the social implications embedded in the language. In such cases, people can support a given power structure while explicitly denying that they are necessarily doing so. They do not use the specific words, but their listeners read racist meanings into them. Because routine forms of accepted discourse generally express continuous but opaque power relationships, they can also be especially useful to particular groups seeking to convey particular social messages while avoiding direct confrontations.5

Through an analysis of the social meanings embedded in the discursive practices of the societal and political institutions, the analyst can thus investigate the nature and reproduction of the power structure. At the ‘macro’ level of analysis, the investigator is able to focus on how political and economic elites construct and maintain a societal‐wide hegemonic discourse that makes clear what is on the agenda and what is not, as well as how oppositional groups seek to make their voices heard. Today we can easily understand this in terms of the construction of a market‐oriented neo‐liberal ideology that now dominates not only Western industrial countries, but much of global discourse as well. The success of such a hegemonic discourse is in significant part measured in terms of the difficulties it poses for those who argue for government interventions in both the economy and the provision of social redistributive programmes. At the ‘micro’ level of analysis, the focus turns to the relationships between discourse and specific institutional practices. The goal, following Foucault, is to undercover the ways that discourses embedded in institutional practices function to reproduce the existing power‐structure relationships from the bottom up. Here, through the workings of professional expertise, ideas and knowledge can serve as an independent force stabilizing or cementing society together. In terms of neoliberalism this can take the form of cost‐benefit oriented policy analysis in agency (p.90) bureaucracies, which implicitly insures the primacy of market criteria, over in preference to social agency practices that emphasize ‘workfare’ counselling. One the surface, cost‐benefit analysis looks like—and is presented as—a scientific technique for rational data collection and decision‐making. But most people do not see that the data collected by the methodology channels discussion in directions biased toward a market orientation of society (Fischer 1987: 117).

Emphasis on institutional practices also helps to clear up the widespread misunderstanding that such argumentative research just deals with languages and, in the process, loses connection with the ‘hard’ institutional realities of policymaking. To the contrary, for the kind of discourse analysis advanced by Fairclough and Hajer, understanding what the relevant utterances mean requires approaching the institutional setting as an ‘argumentative field’ in which statements are made. The focus is thus not simply on the language or discursive constructs but on the ways utterances relate to the specific institutional contexts and practices in which they can be meaningfully stated and understood. Indeed, Hajer (1995b) takes the point a step further and argues that we need engage in ‘institutional constructivism’. This involves analysing which sorts of practices actually produce or facilitate particular policy changes. In the acid rain case, for example, one of his most interesting discoveries is that, after the environmentalists had succeeded in publicly winning the policy debate—at which point everybody began speaking the environmental language, or a version of it—no policy change could be detected. Why? Because the technically oriented practices of the environmental agencies remained unchanged. Although the public political discourse had accepted the environmental understanding of acid rain, the environmental ministries continued to monitor the problem with their traditional methodologies (Hajer 1993). The practices themselves determined the policy, despite the new environmental discourse.


Discourse, as we have seen in this chapter, is an ensemble of ideas and concepts that give social meaning to social and physical relationships. Operating at both the macro and micro levels of society, discourse both transmits the cultural traditions of society and mediates everyday social and political interactions. Focusing on the relationship between discourse and social practices, the task of the discursive analyst is to explain how specific discourses become hegemonic, explicate the characteristics of discursive fields (including the nodal points that privilege some arguments over others), identify the defining claims of the particular positions, clarify how individual discourses come to influence others, determine the structure of the arguments, identify which styles of discourse make them effective in given contexts, uncover the ways that the discursive resources are distributed across social systems, and show how particular socio‐historical constellations serve to justify specific courses of action. Emphasizing the ways discursive production socially constructs reality, such analysis examines the way hegemonic conceptions of reality are upheld and reproduced by key political groups, while oppositional groups seek argumentative (p.91) strategies to challenge these dominant social constructs. Although discourses take place in pre‐structured contexts that limit or impede the range of possible actions, persuasive discourses can even in the face entrenched social and material forces open new paths to action.

More specifically, we saw the ways discourses establish the terrain on which political struggle takes place. Assigning subject‐positions to social actors in both cultural narratives and ongoing storylines, they define the actors' social and political relationships, attributing to them various social attributes such as virtue or blame. In argumentation at the micro or practical level, participants exchange competing and sometimes contradictory arguments to make sense of social reality, including the interpretation of political conflicts themselves. Operating in a field of existing narrative discourses, social and political actors are seen to choose and adapt thoughts and ideas in an ongoing effort to triumph over rival arguments.

The nature of these discursive interactions is further explored in the next chapter, which presents a comparison of Sabatier's theory of advocacy coalitions with Hajer's theory of discourse course coalitions. Of particular importance, the comparison offers the opportunity to take a closer look at the essential role of discourse and interpretation in understanding the processes of social change. Where the former seeks to advance an empiricist theory of policy change, the latter details the postempiricist discursive alternative. The two approaches are compared in terms of a concrete policy issue, environmental protection.

Finally, it should be noted that most of the points in the discussion were made or illustrated through policies pertaining to environmental politics. In significant part, this is because the fields of environmental and science studies, given their close relationship to social constructionist perspectives, have generated a great deal of the research about policy discourse. But this should not be taken to limit the value of the perspective. In the expert‐oriented information society in which we live, the role of discursive politics in policymaking is expected to become increasingly apparent in other policy areas as well. If, as some have argued, the struggle between those with and those without knowledge is one of the key socio‐political dynamics of this new century, the discourses of knowledge in their various forms will be a central aspect of political conflict across the policy spectrum.

Appendix Discourse and Social Change: Commodifying Educational Policy

Fairclough's (1992: 207–15) analysis of educational commodification begins through a textual analysis illustrating the way particular words and images are mixed together to create a social meaning. Towards this end, Fairclough focuses on the how the word ‘skill’ is employed in a university catalogue designed to attract and inform students. The text is structured to implicitly—if not explicitly—inform the students that they need skills to get ahead in their lives and that the educational institution in question is well prepared to supply them with the appropriate skills.

(p.92) After having analysed the linguistic presentation of the text, Fairclough then turns to the analysis of discursive practices to show the way the word ‘skill’ is given a double meaning. By strategically overlaying a term of business discourse on an educational discourse, educational planners and curriculum designers create a rhetoric of skills that can be interpreted in more than one way. As a commodified educational discourse becomes ‘dominated by a vocabulary of skills, including not only the word “skill”, but a whole wording of the processes of learning and teaching based upon concepts of skill, skill training, use of skills, transfer of skills, and so forth’, the concept permits ‘two often contradictory constructions of the learner to coexist without manifest inconsistency, because it seems to fit into either an individual view of learning, or an objectivist view of training’. In addition to the semantic history of the word, as Fairclough writes, ‘this ambivalence is reflected in the history of the concept within liberal humanist and conservative economic discourse’. The concept of skill, on the one hand, ‘has active and individualist implications: skills are prized attributes of individuals, individuals differ in types and degree of skill, and it is open to each individual to refine skills or add new ones’. The concept of skill, on the other hand, ‘has normative, passive and objectifying implications: all individuals acquire elements from a common social repertoire of skills, via normalized training procedures, and skills are assumed to be transferable across context, occasions, and users, in a way which leaves little space for individuality’.

Drawing on laissez‐faire, market‐oriented economic discourse, ‘the message to teachers and course planners is a subtle version of the marketing axiom of consumer sovereignty: “Give the customers what they want’ ”. This juxtaposing of ‘wordings facilitates the metaphorical transfer of the vocabulary of commodities and markets into the educational order of discourse’. But, at the same time, ‘this commodified educational discourse is commonly more self‐contradictory that this might suggest’. A clue to such contradictions in the blurring of clients and consumers is an ambiguity about to whom educational packages are being sold. Who is doing the learning? Is it the students or the companies that are likely to hire the students? The company may in fact need clients ‘in the direct sense of paying the learner to take a course’. In this way, ‘learners are contradictorily constructed’. They are constructed ‘in the active role of discerning customers or consumers aware of their “needs” and able to select courses which meet their needs’. But they are constructed, at the same time, ‘in the passive role of elements or instruments in the production processes, targeted to train in required “skills” or “competences”, courses designed around precise “attainment targets” and culminating in “profiles” of learners, both of which are specified in terms of quite precise skills’. This subtle intermingling of active and passive constructions of student learners ‘facilitates the manipulation of the learners through education, by overlying it with what one might call an individualist and consumeristic rhetoric’.

From the perspective of social practices the university has conflated these separate discourses in such a way that, while they benefit different participants differently, they gradually come to be taken as the same thing. That is, obtaining skill comes to be seen as the essence of education, although the curriculum for these skills is no longer shaped around the liberal arts components, but rather the needs of industry. Instead of the traditional understanding of education as a way for the individual to broaden his or her relationship to the world in which he or she lives, thus opening up new possibilities, it is in fact narrowed down to a package that is shaped to meet the needs of business and industry rather than the individual per se. This is not to say that the individual cannot gain from this restructuring, but it is to say that the choices are dramatically limited to those that also fit the interests of the employers. Subtly, though, in the process the two come to be conflated. The university tries to say at the same time, ‘come to us and get a well‐rounded education’, but at the same time it is saying ‘come (p.93) to us and get a job’. The two are not the same and the tension has to be mediated. This happens through the concept of skill, which blurs the tension in subtle ways. In the process, the university gets restructured. It becomes more of a training centre for industry, with the business school more and more emerging as the primary faculty, rather than a traditional centre for knowledge and learning for itself. Whereas the traditional university was administered by the faculty, as they possessed the knowledge of the curriculum and could best decide the issues of learning, today the deans become managers and the faculty members are increasingly their subordinates. The facade of faculty governance remains, but it increasingly serves to hide the real nature of institutional power and control.


(1) No contemporary German would deny that an emphasis on order and efficiency is embedded in Germany's institutional culture, but many complain that it is not always that easy to find in practice.

(2) Initially, the concept of political discourse was introduced to understand the historical evolution of social systems by analysing the nature, origins, and evolution of its discursive formations. In particular, it has been developed to escape the problematic aspects of the concept of ideology, especially its relationship to the Marxist concept of ‘false consciousness’ and the limits of more materialist approaches that assigned cultural and other ideational dimensions to a secondary status. In the process, though, this relationship of discourse to ideology has become a complicated theoretical question.

(3) The point is well illustrated by the Green Party in Germany. The party gained political influence by joining with the Social Democratic Party to form the governing coalition in 1998. But the Greens' position as junior member of the coalition caused the party to soften its more critical political positions, such as opposition to military engagements, at considerable cost. In response, many party loyalists resigned their memberships, charging the leadership with having politically sold out in the interest of gaining power. In the process, the party created political identity problems for itself that have led many voters to see the party as a superfluous feature of the German political landscape.

(4) The concept of sustainable development was advanced by the Bruntland Commission and was the focus of discussions at the second international Earth Summit in Rio de Janeiro. The Commission defined sustainable development as a ‘framework for the integration of environmental policies and development strategies—the term “development” being used in the broadest sense of the term’. Sustainable development is designed to meet ‘the needs of the present without compromising the ability of future generations to meet their own needs’ (World Commission on Environment and Development 1987: 42).

(5) A good example here is the political rhetoric of Joerg Haider, the right‐wing Austrian politician, who uses words employed by the Nazis. The words say nothing about Nazis but they serve to recall specific memories of a particular political past.