1Roughly two decades ago, a new set of approaches to public policy analysis – which can be loosely grouped together under the umbrella term of “discursive” approaches [1] – emerged in both Europe and the United States. Although these approaches were initially espoused by a few researchers seeking to distance themselves from rational choice theory (the prevailing theory of the time), in recent years the popularity of discursive approaches has grown considerably, to the extent that they are now recognized as one of the major approaches in public policy analysis in most European and US textbooks. [2] While they may subscribe to different premises and even bear different names, all discursive approaches reject the dominance of rational choice theory and, more importantly, share a central analytical focus : discourse, language, argumentation, and, more broadly, the question of “meaning in action”. [3]

2Drawing inspiration from the “linguistic turn” in philosophy and social science, discursive approaches are primarily associated with constructivist and qualitative perspectives. In this sense, they can be distinguished from another set of approaches, where the analysis of public policy is understood as the practice of producing technical or academic knowledge. Discursive approaches justifiably condemn the latter for masking the normativity of their analyses behind the façade of objectivity. Discursive approaches, by contrast, pay particular attention to the subjectivity of actors, to the forms of practical knowledge they mobilize, the multiple interpretations they deploy to create meaning, and the particular context that individualizes the situations in which these meanings evolve. Discursive approaches thus both attempt to fully comprehend the production of meaning, and to analyze the processes through which this meaning shapes actions and institutions. [4]

3Discursive approaches refuse to treat discourse as merely one explanatory variable among many others, considering it instead as the key to understanding how actors construct and modify public policies. In this respect, discourse is first and foremost the means by which actors perceive the world, and consequently functions as a way to transform it. Finally, discourse is what actors use to interact with one another, whether in struggle or in the search for common ground. From this perspective, by honing in on discourse and studying the multiple social practices in which it takes place – when arguing, expressing conviction, conducting analyses, defining terms, attempting to persuade, negotiating, contesting, and so forth – these approaches not only provide a different way of understanding the processes of crafting and transforming policy-making, but also offer a different means of investigating the issues of power, legitimacy and governance that are at the foreground of public action.

4While these approaches can be sharply differentiated from many positivist Anglo-Saxon approaches to public policy analysis, including those which attempt to understand transformation processes by using the causal methods of the natural sciences, [5] they share certain traits with French approaches that should be noted. As with cognitive approaches, [6] discursive approaches place the question of the meaning of public policy center stage, and reject the distinction between ideas and interests, [7] believing both to be social constructs. In common with approaches that focus on policy instruments, [8] they often refer to the work of Michel Foucault and the sociology of science, rejecting the notion of the instrument as a neutral object devoid of meaning and effects. In common with pragmatist approaches, [9] they pay serious attention to the production of actors’ practical knowledge, its use, and the way in which knowledge is challenged in general. Furthermore, discursive approaches consider defining both the problem and the solution as essential. More broadly, and in common with much current research in France, discursive approaches address the political dimension of public policy.

5Despite such shared traits, as well as the popularity of discursive approaches elsewhere in Europe, [10] they remain relatively unknown in France. The importance of an applied dimension in many discursive approaches – with the avowed purpose of improving democracy – may partially explain this lack of recognition. Although it may clash with many French approaches, an applied dimension – distinctly present in much Anglo-Saxon public policy research – should not prevent us from examining the ways in which these discursive approaches enrich our understanding of processes. By focusing on the significance of the production of empirical and analytic knowledge, as well as on the often hidden normative dimension of these forms of knowledge, and the ways in which they are challenged through discursive interaction, discursive approaches force analysts to rethink the role of knowledge in processes, and to question the supposedly rigid barrier between empirical knowledge and normative approaches.

6The goal of this special issue is thus not only to present discursive approaches and to highlight the points of convergence that link them to French research, but above all to emphasize their specific contributions to the policy debate and how they enrich it. With that in mind, we are less concerned here with simply importing these approaches than we are with provoking an international dialogue on the meaning and political dimension of public policies. Instead of attempting to cover the entire field of discursive approaches in this issue, we have instead decided to present three articles in the hopes of prompting this discussion.

7Consequently, we have chosen to exclude both the studies best known to French researchers – particularly those dealing with narratives [11] – and those works primarily focused on the production of new analytical methods designed to solve social problems. [12] The articles presented here are instead critical studies on public policies ; that is, they attempt to observe discursive practices as a way of better understanding not only policy-making processes, but also more broadly the forms of governance and the political stakes underpinning them.

8Before presenting these articles and highlighting their contribution to the field of discursive approaches, we shall first look at how these approaches emerged, what conditions made their emergence possible, and the researchers who first espoused them ; moreover, via a certain number of key references, we shall outline the hypotheses shared by discursive approaches, as well as the differences to which they lay claim.

The emergence of a critical stance towards and within public policy analysis

9In order to better describe discursive approaches, we shall begin by referring to the key studies that inspired them, as well as the field’s earliest works. Rather than compile a long and inevitably incomplete catalogue of works, we have chosen instead to concentrate on two major influences that directly inspired the first studies claiming to employ a discursive approach – namely, Jürgen Habermas and Michel Foucault – before addressing the most important of the early studies on the discursive analysis of public policy.

10Habermas’s theory of communicative action [13] laid the groundwork for discursive approaches both through its critique of technological rationality and the notion of communicative reason. [14] Proponents of discursive approaches took up Habermas’s critique of the use of rationality in science and technology, which he saw as rooted in ideological knowledge and partisanship rather than in objective and apolitical knowledge. [15] Discursive approaches suggest applying the same style of critique to the technical analysis of public policy, in order to better highlight that what appears technical is in fact concealed normativity. In the same manner, discursive approaches employ Habermas’s notion of communicative rationality to illustrate the importance of concrete discursive interactions between actors as the production sites of discourse and meaning.

11From Surveiller et punir, Les Mots et les choses to L’ordre du discours, [16] the works of Michel Foucault [17] are, alongside those of Habermas, undoubtedly the most frequently cited in discursive studies. Foucault’s reflections on the role of discourse as the site of the production of knowledge and power were the first to be widely acknowledged by public policy researchers outside of France. The work of Foucault advanced a better understanding of the importance of discourse as the site where a compromise – both incidental and contingent – could be reached between the normal and the pathological, the true and the false, the permitted and the prohibited, between reason and madness. He highlighted the forms of power which these discourses contain but also permit. For Foucault, “discourse is not only that which translates struggles and systems of domination, but it is also that for and through which we struggle, the power we seek to make ours’. [18] Drawing on his work, researchers have tackled not only the production and use of discourse but also how it is mastered, plotted, disseminated or excluded, which are all instances where the question of power is played out. [19]

12From the end of the 1980s, political scientists began to address the issue of discourse in public policy. Drawing on the linguistic turn in the field of social science, they attempted to apply it to policy analysis. We shall now review the work of several authors who played an essential role in this process of theoretical importation and consequently in the emergence of discursive approaches.

13The first two authors to spark debate over public policy analysis were Deborah Stone, who in 1988 published Policy Paradox, [20] and Giandomenico Majone, who published Evidence, Argument, and Persuasion in the Policy Process a year later. [21] In their studies, both argue that it is impossible to analyze public policies objectively and rationally, and they further illustrate that all policies are first and foremost discursive constructions [22] that combine heterogeneous elements such as values, instruments and consequences. Consequently, rather than using rationality to analyze public policies, these authors suggested studying actors’ analytical production as well as their use of arguments. Rather than searching for the objective “meaning” of a public policy, Stone and Majone proposed studying how actors interpret policy, and the interpretive struggles that these actors wage in order to influence the choice of policy instruments. For example, they examined how arguments regarding feasibility can circumscribe the field of possibilities by imposing limits.

14In 1993, Frank Fischer and John Forester edited the volume entitled The Argumentative Turn in Policy Analysis and Planning which brought together the work of a dozen or so authors for the first time. [23] As a result, researchers that had hitherto been working independently were finally able to participate in a truly collective endeavor. [24] Exploring issues as diverse as think-tanks in the United States, acid rain in the United Kingdom, healthcare reform in Oregon and electricity policies in Chicago, the contributors to the volume shed light on the constructed, normative and political nature of the knowledge produced by policy actors, and the significant impact of their actions on public policy development and implementation. [25] By illustrating the relative and political nature of all analytical production, these authors argued that public policy analysis should no longer be limited to highlighting the irrationality of certain policies, [26] but should rather re-establish politics center stage. [27]

15The number of publications addressing the question of discourse in public policy continued to increase throughout the 1990s. A brief overview would include the work of Maarten Hajer on the struggles between discursive coalitions over environmental policy ; [28] Dvora Yanow [29] on new methodologies enabling researchers to better account for the plurality of interpretations of problems and categories of public action ; Steven Griggs and David Howarth [30] on aviation policy ; Herbert Gottweis on biotechnology policy ; [31] Mark Bevir and Rod Rhodes on governance narratives and the everyday activity of government ; [32] and Vivien Schmidt [33] on discursive institutionalism.

Discourse as the focus of public policy analysis and process studies

16In order to contextualize this rather varied and heterogeneous shift in policy analysis, let us refer to two guiding principles which structure many works adopting a discursive approach. This will allow us to better illustrate the value of an approach which places discourse at the heart of its analysis.

17First and foremost, these works are grounded in an approach where discourse is not simply the reflection of what actors call “reality”, reduced to a mere problem of interpretation, but also the means through which they can shape this reality. Language is therefore not only a tool that enables interaction, but also an instrument of power. [34] Discursive practices are even more interesting to observe because they not only allow us to describe certain positions, but also to redefine them. Changing one’s position, persuading certain groups, excluding others, defining legitimate authority and establishing dominant groups are all processes that play a part in the distribution of positions in the political arena. [35]

18Drawing inspiration in equal parts from constructivist theories of knowledge, Wittgenstein’s “language games” [Sprachspiele], [36] and Austin’s speech acts, [37] authors using discursive approaches emphasized both the always-partial and biased cognitive frameworks that actors must use to apprehend reality, and the ways in which these frameworks transform reality. When applied to public problems, this first guiding principle can be observed in numerous public policy analyses – and not only those employing a discursive approach. Since Harold Lasswell, Charles Jones, Peter Bachrach and Morton Baratz, up to and including contemporary authors such as Frank Baumgartner, Wayne Parsons and Renate Mayntz, [38] it has been widely accepted that the possibility of an issue landing on a political agenda is not tied to its intrinsic or objective value, but rather to the ability of policy “actors to define, promote, and transform it into a serious public issue. Analysis through the lens of discourse, then, makes possible both the problematization of situations and their politicization, through an observation of the conditions that make them public issues”. [39] This discursive politicization of issues thus depends on the meaning attributed to them by competing social groups. It particularly depends on the ability to discursively incorporate the causes of the problem, those accountable for it, and the public authorities who are called upon to solve it.

19If adepts of the discursive approach are working in the long tradition of the sociology of social problems, their principal contribution has been to demonstrate that this guiding principle can just as easily be applied to policy solutions as to the identification of policy issues. As Stone has suggested, the “solutions”, far from being neutral instruments whose meaning and dynamics can be objectively established, are in fact constructs to which actors and groups ascribe meaning, values, consequences and the ability to resolve problems. [40]

20The second guiding principle of these studies is that discourses can only be understood through the social practices in which they are embedded. While the concept of discourse encompasses ideas, representations, forms of knowledge, values, beliefs and norms, it also inextricably links these elements to the individuals who are speaking. Discursive practices are therefore always understood reflexively as interactions, in which many elements are played out at once : an exchange of content, but also of intentions, identities, positions and power. In this respect, this approach follows in the pragmatist tradition which spans John Dewey and George H. Mead, via Harold Garfinkel, [41] and its proponents thereby position themselves as theoreticians of practical knowledge. [42] According to this theory, rather than merely being the reflection of ideas, social practices help to fashion and transform them.

21These authors therefore focus on specific discursive practices such as argumentation, persuasion, negotiation, conviction, definition, comparison and injunction. These social practices consistently involve actors who possess specific intentions and discourses, and who interact with other actors, against whom they test these discourses and intentions. Generally speaking, researchers differentiate intentions from their effects in order to illustrate that intentions in no way imply effects, and can in fact even produce unintended outcomes. This is what Austin refers to as the illocutionary and perlocutionary effects of discourse. Although the intention is often to achieve agreement and assent, the effect produced can instead be one of indifference or opposition.

Three texts for rethinking the political dimension of public policies

22This special issue comprises three articles, each of which not only represents a different trend in the field of discursive approaches, but which also takes a specific approach to the political dimension of public policies.

23The first article deals with the “argumentative” turn, developed by Frank Fischer in collaboration with John Forester, and which finds its most well developed expression in Fischer’s Reframing Public Policy. [43] The argumentative approach emphasises the critical dimension of the activity of the researcher, who should shed light on the power stakes and the normative framework concealed behind policy arguments. Argumentative analyses should seek to describe what is at stake in a given situation, and therefore contain both a descriptive dimension that examines normative strategies, and a prescriptive dimension that identifies discussion as a means of highlighting both the normative frameworks of arguments, and the way in which they clash with one another in the search for agreement.

24In the second text, drawing primarily from hermeneutic philosophy, Mark Bevir presents the “interpretive” approach, which focuses on the irreducible plurality of interpretations. Firmly situated within the field of political theory, Bevir’s work thus attempts to simultaneously account for actors’ intentionality, historicity and reflexivity in order to better comprehend the forms of governance in play. Following his collaboration with Rod Rhodes, Bevir shows that it is necessary to go beyond the public policy network approach and incorporate narrative elements and storytelling into the analysis.

25The third article, by one of the present authors Philippe Zittoun, presents a pragmatic approach to public policy. It examines how policy proposals are constructed, paying particular attention to three elements : the normative construction of knowledge wherein policy proposals take shape ; attempts at giving these proposals meaning within coherent statements ; and the way in which these proposals are challenged during the discursive interactions that occur throughout the decision-making process. By observing “discourse in action”, this article goes beyond the opposition between a sociological approach to actors that ignores ideas and a cognitive approach to public policy that all too often ignores actors.

26While discursive approaches aim to understand and describe the multiple productions of meaning, they cannot avoid the thorny issue that they themselves must produce meaning in order to reach such an understanding. Discursive approaches often reflect the difficulty of using language as an instrument of analysis, given that they qualify language as ambiguous and difficult to pin down. In this light, it is of no small significance that one of the articles in this issue – and, indeed, this introduction – was written in French but is published here in translated form. [44] The translation has thus had to alternate between two tendencies : on the one hand, identically reproducing the – often abstract and prolix – French style ; and on the other hand, altering the French original in order to make it more approachable for an Anglo-Saxon readership, at the risk of deviating from the author’s original intentions. We present this translation as a work-in-progress that illustrates what is at stake in an approach based on the social, contingent, complex and dynamic nature of language.