Locating and Applying Critical Discourse Analysis within Education Policy : An Introduction

This article introduces the first of a two-part Special Issue on Discourse Perspectives and Education Policy. This first special issue is focused on critical discourse analysis and education policy. Within this article, we provide a brief overview of discourse analysis generally and critical discourse analysis specifically. We highlight some of the ways in which policy researchers have applied the theories and methods associated with CDA and note the methodological and substantive contributions of this work. Then, we provide Critical discourse analysis within education policy 2 an overview of the six papers included within this special issue, noting each paper’s key points and explicit links to policy. We conclude by pointing to future directions for research at the intersection of education policy and discourse studies.


Introduction
Education policy scholars have given increasing attention to the competing issues of justice and injustice, equity and inequity, as well as policy issues relating to specific populations (e.g., students with lower socioeconomic status, special education eligible students, students who do not speak English as their primary language, LGBTQ students, among others).Within this attention, scholars have advocated for the use of critical perspectives and frameworks (e.g., Diem, Young, Welton, Mansfield, & Lee, 2014;Young & Diem, 2014).This advocacy has shifted the policy discourse considerably toward studies which essentially ask: whose interests does the policy serve?This shift can be noted methodologically, as policy scholars have increasingly turned to discourse analysis approaches, such as critical discourse analysis (CDA), to examine underlying issues of power, social inequality, (re)production of dominance, as well as the formation of identities within policy 'talk' and/or legislative texts.This examination is proving fruitful on a number of levels, as the papers in this special issue demonstrate.For example, the papers included in this first of a two-part special issue broadly focused on discourse perspectives and education policy, focus on a range of policy issues and attend to concerns at the state level and those at the local level (i.e., within particular schools and districts), as well as policy issues relating to both policy implementers and policy actors.This broad focus is noteworthy as it demonstrates the utility of CDA to consider both micro-analytic (e.g., how policy implementers orient to and make sense of policy within a given organizational context) and macro-analytic (e.g., how formalized actors espouse and reproduce particular ideologies) concerns.Prior to examining how each of the contributor's to this special issue take up CDA and attend to policy-related issues, we provide an overview of discourse analysis generally and CDA more specifically.Then, we discuss each of the papers included in this special issue, identifying their key points and explicit links to policy.Finally, we conclude by noting future directions for the fields of education policy and discourse studies.Indeed, we see these fields as being integrally connected and thus use this special issue to highlight and advance the substantive and methodological connections.

An Abbreviated Overview of Discourse Analysis
The field of discourse studies emerged in the 1960s as various forms of discourse analysis were developed and increasingly applied within the social sciences (Angermuller, Maingueneau, & Wodak, 2014).As a field, discourse studies has been influenced by a range of disciplines (e.g., anthropology, linguistics, philosophy, sociology, among others), intellectual traditions (e.g., interactionism, Marxism, poststructuralism, pragmatics, among others), and scholars (e.g., Derrida, Foucault, Goffman, Habermas, Sacks, among many others), and has been described as "one which runs counter to the division of knowledge into specialized disciplines and sub-disciplines" (Angermuller et al., 2014, p. 1).Within such a diverse field, then, it is perhaps unsurprising that there is no single definition of discourse or single approach to carrying out a discourse analysis.Further, even within one tradition of discourse analysis, such as CDA, analysts often take up a diverse range of approaches to studying language.
Generally, it is helpful to think about 'discourse analysis' as an umbrella term used to describe an interdisciplinary set of qualitative approaches used to study talk and text in social life (Potter, 1997).While the landscape of the theories and methods related to discourse analysis is vast and quite diverse, across the spectrum of approaches, researchers generally argue that talk and texts do not neutrally reflect the world, one's identity, or social relations (Howarth, 2000).Thus, it is assumed that meaning is not a given; rather, meaning of all kinds is produced in and through social practices.The discourse analyst, then, presumes that language is performative; that is, it is always doing something with consequence (whether intended or not).For instance, it is through language that one goes about constructing their identity, ascribing identities to others, positioning others, etc.These ideas certainly predate discourse studies and discourse analysis, as they can be traced back to linguistic philosophers such as Wittgenstein (1958) and Winch (1967), to name a few.Further, many discourse analysts also presume that it is through language (broadly defined) that inequalities are (re)produced and sustained, particularly as taken-for-granted discourses and practices become naturalized.Indeed, it is this very 'taken-for-granted-ness' that many critically informed discourse analysts seek to unsettle and deconstruct.
More particularly, within discourse analytic traditions, a researcher may study the function of language at a micro (e.g., conversational, see Gabriel & Paulus, 2015), meso (e.g., institutional, see Anderson, 2015), and/or macro (e.g., social or historical, see Peters, 2007) levels.They may accordingly take a more pragmatic view of language (e.g., Gumperz, 1982) or one that it is steeped in a socio-historical perspective of discourse (e.g., Foucault, 1989), among other positions.In policy, such a discourse analytic approach may allow for the study of everyday interactions; for instance, an analyst could examine how teachers construct (through their talk) an understanding of accountability expectations within the context of their teaching practices, or the ways in which school leaders articulate conflicting constructions or descriptions of learning improvement needs across organizational contexts.One might also study how a state level policy has historically (re)produced a particular version(s) of what it means to be a 'successful' or 'ideal' student.

A Brief Overview of Critical Discourse Analysis
Within the diverse landscape of approaches to the study of language, CDA has been frequently drawn upon in policy related studies to consider issues related to power, inequality, ideology, and "the exercise of social power by elites, institutions or groups, that result in social inequality" (van Dijk, 2001, p. 249).As van Dijk (2001) argued, CDA is "primarily interested and motivated by pressing social issues, which it hopes to better understand through discourse analysis" (p.252).Fairclough suggests that it is "the analysis of linguistic and semiotic aspects of social processes and problems" (p.368).While CDA seems to describe one framework for discourse analysis, in practice CDA represents a range of approaches to studying language (e.g., Fairclough's dialectical theory of discourse, van Dijk's cognitive approaches to CDA, Wodak and colleague's discourse-historical approach, among others).Unsurprisingly, then, this discourse analytic approach cannot be traced back to one origin (Jørgensen & Phillips, 2002); rather, it is broadly informed by critical theories and Foucault's perspective on power as productive (Foucault, 1990).Like many discourse analysts, the CDA scholar rejects the idea of "value-free…objective" research and in contrast often positions a CDA project as intervention-oriented and in alignment with the positions of the marginalized (Fairclough, Mulderrig, & Wodak, 2011, p. 358).In other words, the project itself is decidedly ideological with the scholar positioned as part of the very discourse being critiqued ( van Dijk, 2001).While historically CDA has focused on macro-level discourses, it also has served to link micro level practices (i.e., everyday interactions) with macro concerns (i.e., the broader historically sociopolitical context).
In contrast to some of the other approaches to discourse analysis, CDA has often been described as taking a more 'top down' approach, which presupposes the relevance of issues such as power and inequity.This perspective has certainly not been without critique, as some scholars have questioned the use of a priori categories (e.g., "race") being made relevant by analysts at the start of a project rather than emerging as relevant within the interactions of social actors under investigation (Benwell & Stokoe, 2006).

Education Policy and Critical Discourse Analysis
While policy is defined broadly (see, for example, Ball, 2006), many definitions do not preclude the use of discourse perspectives to understand and analyze policy and everyday practices related to it.As Cibulka (1995) noted, "Policy studies actually comprises a number of separate subfields with distinct, although overlapping, emphases" (p.107).Therefore, at the level of methodology, we argue that the use of discourse perspectives may be essential in helping the field understand substantive policy matters given the breadth of disciplines upon which policy research tends to draw.Across these many subfields, there are a growing number of examples of discourse methodologies being employed, including political science (e.g., Bhatia, 2006;Townshend, 2004), sociology (e.g., Perez, 2013), and educational psychology (e.g., Lester & Gabriel, 2014), among many others.
More specifically, CDA has been widely used in education and education policy (Rogers et al., 2016).While our discussion here is necessarily brief, we direct the reader to a comprehensive literature review carried out by Rogers et al. (2005Rogers et al. ( , 2016)), which provides a critical synthesis of CDA in education research and a useful review of the history of CDA in education research.In education policy, analysts have used CDA to investigate both the substance and role of language in each phase in the life cycle of a given policy from the construction of a policy problem, through varied aspects of the policy-making process, to the reaction, representation, implementation and critique of existing policies.In each case, CDA has been used to analyze the talk and texts that inform and constitute the social realities and understandings of policy related to education.This has involved the analysis of everything from mission statements and textbooks produced in response to a policy mandate (Ayers, 2005;Barton & Sakwa, 2011), to transcripts of meetings and evidence considered during policymaking processes (Arnott & Ogza, 2010;Gabriel & Lester, 2014;Scollon, 2010), to name only a few examples.Outside of policy meetings and policy documents, analysts have also considered the representations of policy and policy problems in the media, the responses of educators and other stakeholders, and the text and talk that surround implementation and evaluation efforts over time (Cohen, 2010;Gabriel & Lester 2013;Kuntze, Gildersleeve & Pasque, 2011).Notably, CDA offers a comprehensive toolkit with a wide range of possibilities related to education policy research that have, at their core, an understanding that language is both the content and conduit of policy.Thus studies of language can illuminate not only the substance or impact of education policy, but also the processes by which such substance and impact come to be, and come to be understood.

Overview of the Papers Included in the Special Issue
The six papers included in this first installment of two special issues devoted to discourse analysis in policy studies represent a range of methodological decisions and combinations within CDA.More importantly, however, they demonstrate some of the diversity of possibilities for what counts as data and what counts as policy within discursive investigations of educational policy.Though some authors investigate policy as a noun, by analyzing the "letter of the laws" themselves, the discourses of policymakers, and media representations policies as they are constructed and ratified; others investigate policy as action: the discourses of policy actors as they construct, implement, resist and transform education policy.
Hernandez examines how reformers and politicians legitimate their claims about educational problems and solutions, particularly investigating how language about race complicates the persistent claim that education reform is a civil rights issue.By examining the constructions of race within interviews including a prominent policymaker and a prominent "reformer," Hernandez demonstrates patterns in the construction of race within civil rights rhetoric and explains the social, historical and political significance of these constructions as they relate to education policy and reform efforts.In this case, a micro-level analysis, informed by theories of language use, critical race theory and an understanding of policy as the positioned instantiation of ideology produces an analysis that explains and critiques a pattern of conflict and failure within the discourse of education reform.
Hernandez's investigation is unique in that it uses both researcher-generated and publically available interviews to juxtapose the language use by two influential actors in the education reform arena.In doing so, she compares both artifacts of research (1:1 interview) and media representations (e.g.televised interviews) to identify patterns in the pervasive rhetoric of civil rights within education reform both in public and private conversations.In this case, CDA is informed not only by understandings of the broader social and political phenomena under investigation, but also by an understanding of the function of particular linguistic devices used when talking about race and racial ideology.
Emery also used interviews with policy actors to investigate the relationship between language and ideology, but his application of this approach in a different arena and using different theoretical perspectives and methodological tools demonstrates some of the range of possibilities for such approaches.Emery analyzes the discourses of social and emotional learning in schools as understood and enacted by policy makers in England and Wales during the New Labour years (1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010).Situating this study in the context of a devolved government across the UK, during a particular political moment in which neoliberalism constituted a hegemonic "common sense" discourse, Emery investigates how policies related to social and emotional learning were framed by policy makers and how this framing was influenced by certain understandings of childhood wellbeing, tradition, history and identity.In doing so, Emery puts interviews with individual policymakers in conversation with each other, identifying points of interdiscursivity -where participants talk connects with wider social practices and ideas.
By tracing the history, association, and connection of the discourses participants draw upon to talk about social and emotional learning, Emery isolates the ingredients that combined to construct particular ideologically-driven versions of social emotional learning in two countries (England and Wales) in this particular historical moment.Illuminating the mix of ideas that constitute taken-for-granted assumptions in both countries, Emery creates space both to critique and offer alternatives.This demonstrates how CDA, and in particular Fairclough's three dimensional model (1992), can be applied in ways that do not only uncover what is there, but may also drive change.
The concepts of intertextuality and interdiscursivity in Emery's piece can also be traced through a very different application in a study of restrictive language policies in U.S. public schools.In their paper, Jimenez-Silva, Bernstein, and Baca investigate how state policy is interpreted by three contrasting school districts by examining public statements about the policy on state and individual district websites.In this case, public statements of and about policy are taken as data as the contexts of different districts, and indeed the role of context, are analyzed between texts and over time.This paper presents a detailed investigation of ideas that are central to discursive approaches to policy analysis: that policies are constructed by and constructive of discourses, that tracing the history, development, (re)combination and contrasts of discourses used to write or talk about policy illuminates the role of contexts as well as the stakes, interests and ideologies involved.In this case, differences in three districts' interpretations of one state policy are used both to explain variations in implementation across districts and to identify ideologies in context.The authors argue that the latter is particularly important for teachers considering positions in different districts as well as parents or other stakeholders who may benefit from understanding the particular position of a district in relation to a state policy.In this way, discourse analysis of publicly available policy statements do more than just explain patterns of implementation, they may also be used to inform decision-making where local language ideologies are a topic of concern.
Jimenez-Silva et al. investigate policy documents related to the controversial issue of language policy at the district level where policies are translated and filtered from state policy makers to district administrators and finally to school-level actors.However, controversial policy documents can also be examined using the tools of CDA before policies are even constructed.For example, Little and Cohen-Vogel analyzed documents drawn from the websites of organizations representing different perspectives in the debate over kindergarten (developmentalists versus academic advocates).
Little and Cohen-Vogel use CDA to understand similarities and differences in how language is used to construct the purpose and possibility of kindergarten within documents published by competing advocacy groups.Though these groups have opposite ideas about kindergarten as it is, they share common goals for kindergarten, which the authors uncover by applying a causal stories framework (Stone, 1989) to analyze the problem narratives constructed by each advocacy group's online documents.Unlike the first three papers, which analyze the nature and impact of existing policies, Little and Cohen-Vogel focus their analysis on documents that are used to inform policy, thus uncovering the discursive resources available to policy-makers and lobbyists during the policymaking process.Such work is particularly important in arenas where perennially controversial issues are frequently the focus of legislation (e.g.kindergarten policy).In this case, understanding the resources that inform new policies highlights common ground between polarized groups, and clarifies the substantive differences that are too often exaggerated and obscured by assumed ideological divides.
In contrast to studies that investigate discursive resources for policy-making or the impact of existing policies, Matthies examines the process of policy-making while policies are under revision.Matthies created transcripts from audiorecordings of four meetings of a task force assigned to revise a state law regarding the integration of public schools.In order to analyze transcripts of task for proceedings, Matthias applied Fairclough and Fairclough's (2012) approach to discourse as argumentation, working to identify how claims for action were formulated, and how beliefs about the purposes of both public schools and state policy informed competing claims.Her findings can therefore not only be used to understand perspectives on school integration, but also the relationship between school integration policy and underlying epistemic differences regarding beliefs about the purpose of schooling and the role of state policy.
Mining the discourse of policymaking in action allows Matthies to understand what resources and repertoires are included in the discussion and debate that shapes state policy.Unlike the other papers, in which discourse is drawn from documents and interviews, Mattheis investigates language as it occurred in a particular, occasioned setting.Her findings may also inform understandings of the discourse of policymaking processes, and the role of beliefs and values in policymaking conversations.The analysis of naturally occurring discourse is not limited to particular meetings or group processes.However, the consequential constitutive nature of official meetings presents an important site for inquiry into the public construction of policy, which may provide insight into the issues and conflicts that impact implementation.
Lenhoff and Ulmer's investigation of the implementation of programs from one reform organization in a particular district provides another example of discourse analysis applied to the translation of policies into practices.As Lenhoff and Ulmer demonstrate, the discourses of reformers and implementers can be misaligned and in conflict with one another in ways that influence the equity and fidelity of implementation.Lenhoff and Ulmer use a wide range of data sources including field notes, interview transcripts, publically available documents and other source materials collected over a five-year period to understand how a reform organization promoted its program, and how local actors explained the implementation of that program.
Similar to Emery, Lenhoff and Ulmer found important contrasts in the language that different stakeholders use regarding the same policy.In the case of the reform organization's program, local actors questioned whether the program was intended or necessary for all or only some of the students while the reform organization.This misalignment between understandings of the audience for the reform led to uneven implementation and, in some cases, the dismissal of the organization's program on the part of local actors with conflicting views.In this case, understanding how different stakeholders use language to construct misaligned versions of the same program not only explains patterns of implementation, but could be used to inform decisions about program adoption in the first place.As these authors argue, a close analysis of the match between ideologies present in the language used to promote a program and ideologies present in the local context could inform initial selection and ensure greater alignment of program goals and implementation efforts.

Conclusions and Future Directions
As we noted previously, the meaning of policy is often not identified within the letter of a law or the construction and expression of political ideals; rather, it is often shaped and reshaped by policy actors and implementers in conversations that occur at multiple levels of the policy system over time.Within the papers included in this special issue, CDA is applied to a range of data sources taken from the full spectrum of the policy process from problem formation to policy making, translation and implementation.Each paper demonstrates an application of CDA notably in combination with one or more complementary theories and/or methodological frameworks, which shows some of the many possibilities for analyzing discourse and connecting discourse theory and methods with other theoretical and analytic tools from a range of traditions and disciplines.Taken together, these papers highlight not only what can be accomplished by researchers who draw upon a CDA perspective, but also what discourse itself does within the formation, interpretation, and implementation of education policy.
Further, as illustrated by the papers in this special issue, CDA can be applied when a policy scholars to seeks to understand the complexity of policy issues on a number of levels.For example, it can be used when scholars seek to explain and/or critique extant patterns within society, policymaking institutions, and local communities.It often serves as a basis for questioning takenfor-granted assumptions about the particular ways in which policy has come to exist and be known among policy actors, implementers, and societal stakeholders.In doing so, it offers alternative understandings about these issues that can inform decision-making and advance the interests of often marginalized populations.This may be particularly helpful when local language ideologies are a primary topic of concern or when resources are bound within these structures and thus misaligned to the needs of under-served and/or targeted stakeholders.Within these studies, CDA provides a valuable gateway through which policy actors, implementers, and stakeholders can identify new opportunities for common ground and seek to ameliorate differences between hyper-politicized groups by surfacing ideological barriers to collaboration, cooperation, and consensus.This, in turn, enables scholars to assist policy actors in identifying underlying epistemic differences that relate to specific policies or reflect locally derived understandings of how policy should work.Such understandings are often key to breaking down and changing the ways in which policy is implemented and thus CDA might well assist policy actors in promoting greater alignment between program goals and the populations they intend to serve.
While CDA is a powerful tool to understand some important aspects of policy, we also acknowledge its limitations and thereby advocate for continued work at the intersection of discourse studies and education policy.As noted, discourse studies is a broad field and thus there are a multitude of theoretical and methodological possibilities yet to be applied.Indeed, we argue that the applicability of discourse theories and methods to policy is so timely that further research is needed to understand how taking up discourse perspectives might serve to examine, critique, understand, and potentially explain how policies are derived and come to be understood within the context of education practice.We see this work as groundbreaking in that it provides theoretical and methodological substance to an area of policy research that has, until now, fruitfully depended largely on ethnographic and case study methods.As such, we call upon the field to continue exploring the applicability of discourse perspectives to the analysis of education policy, and position this special issue as offering possibilities for the place of CDA in education policy.