ISSUE CONTEMPORARY APPROACHES TO THE STUDY OF EDUCATION POLICY & DISCOURSE education policy analysis archives

: In this article, we introduce the special issue focused on diverse perspectives to discourse analysis for education policy. This article lays the foundation for the special issue by introducing the notion of a third generation of policy research – a strand of policy research we argue is produced at the intersection of education policy and discourse analysis. We also very


Introduction
In a recently published special issue of Education Policy Analysis Archives (Lester, Lochmiller, & Gabriel, 2016), we highlighted the utility of critical discourse analysis (CDA) for attending to policy issues writ large. Notably, CDA has been widely used in education research (Rogers et al., 2016) and could perhaps be characterized as the dominant discourse analytic approach used by education policy scholars (Lester, White, & Lochmiller, in press). Yet, it is important to note that CDA, which encompasses a diverse set of approaches, is located within the broad and diverse landscape of discourse analysis; that is, there are a multitude of theories and analytic perspectives that policy scholars might draw upon when working at the intersection of education policy and discourse analysis. As such, the purpose of this special issue is to contribute further to the scholarly conversation around how diverse approaches to discourse analysis may serve to advance understanding of how policy is constructed, enacted, implemented, and resisted at the level of language (defined in varying ways). Indeed, we are not alone in noting the value of taking up a discourse analytic perspective to study policy, as several scholars have written about the intraconnections between policy and language. For instance, scholars have described policy as text and discourse (Ball, 1993), as well as "policy at the level of language use" (Spolsky, 2004, p. 218).
More particularly, we position the contributions within this special issue in relation to what we are loosely referring to as a third generation of policy research. Over 10 years ago, Honig (2006) described two generations of policy researchone focused on measuring the impact of policies within a more positivist orientation and a second focused on the complexities of policy practices with an epistemological focus on the varying ways that policy is implemented. As Honig noted, "contemporary education policy implementation research can be distinguished epistemologically by its orientation to the nature of knowledge and knowledge-building about implementation" (p. 20). Building upon these ideas further, we suggest that perhaps a third generation of policy research is one that takes up discourse analytic perspectives in varying ways and thereby moves the field to a closer understanding of the varying education discourses and everyday conversational practices that function to create and codify policy institutionally and within specific educational strategies.
Prior to discussing the unique contributions of the six articles included in this special issue, we first provide an abbreviated overview of discourse analysis. Following this, we discuss the individual articles, noting the methodological and substantive contributions. We then point to the collective policy and methodological contributions that this special issue makes, while also noting the possibilities for future research directions.

Discourse Analysis: A Brief Overview
As we have noted elsewhere (Lester et al., 2016), it is perhaps most useful to conceptualize discourse analysis as an umbrella term that includes within it a variety of qualitative approaches to the study of language (defined broadly) (Jørgensen & Phillips, 2002). There is no single definition of discourse analysis or discourse, yet collectively scholars using discourse analytic perspectives focus broadly on studying language as related to social practice (Potter, 2004). Discourse analytic perspectives range from CDA to discursive psychology to Foucauldian discourse analysis to Bakhtinian discourse analysis to interactional sociolinguistics, among others.
Across many of the discourse analytic perspectives there are several shared assumptions. First, language is typically assumed to be performative, as it is understood as the medium by which social life is accomplished. For instance, through language-in-use people negotiate, complain, account for actions, etc. In other words, language is action-oriented. Second and related to the performativity of language, many scholars using discourse analytic perspectives take up a social constructionist position, and thereby position reality and knowledge as constructed in and through language (Burr, 2003). Within this orientation, it is generally assumed that language is not neutral nor a simple reflection of realitya position espoused by linguistic philosophers (e.g., Wittgenstein, 1958). Finally, in differing and varied ways, discourse analysts take up critical understandings of the world as they critique taken-for-granted knowledge and practices.
Despite these common assumptions, since the 1980s the proliferation of discourse analytic approaches has brought with it great variability. This variability includes the very conceptualization of discourse, analytic foci, the preferred data sources (e.g., text -based, video data, computer mediated communication, etc.), and even the way in which analyses unfold and are represented to broader audiences. In many ways, such differences point to the interdisciplinary nature of discourse analysis generally, as it has been influenced by multiple disciplines, theories, and analytic perspectives (Potter & Wetherell, 1987).
Drawing upon our own commitments to work within, across, and even at times against methodological traditions and perspectives, we included in this special issue articles that drew upon diverse approaches to discourse analysis and offered incisive perspectives regarding applications of discourse analysis and discourse theories for policy scholars. Even still, we acknowledge that the articles included in this special issue highlight the diversi ty of discourse analytic perspectives only in part.

Overview of Included Articles
The six articles included in this second installment of two special issues devoted to discourse analysis in education policy represent some of the breadth of the landscape of theories and methods associated with a discourse analytic approach to education policy research. The articles within this special issue draw upon data obtained from a range of sources and attend to policy issues related to contemporary issues. Collectively, these articles are ultimately concerned with a common interest in using analyses of discourse (defined in varying ways) to understand new ways forward for communities, communication, and commitments, particularly as language is used to construct and resist policies in action.
Burman and colleagues apply and stretch the boundaries of CDA in the opening article of this special issue. By examining the subject-formation of those addressed by education policy, the authors examine how participants' narratives demonstrate, but also resist subjectivities made relevant by particular frameworks for thinking and talking about poverty, and specific formulations of people and things as policy problems. In doing so, they argue for a Foucauldian-influenced discursive approach that does not stop at the identification of hegemonic power structures, but pays analytic attention to the action-orientation of the narratives themselves -not merely as reflections of policies and political structures, but as sites of resistance, struggle, and reformation.
In the second article, Wilinski similarly addresses the sharp point of policy -where policy discourses are taken up, shifted and resisted by those who are made the object of policyspecifically, educators responsible for implementing prekindergarten (PreK) policies. It is no coincidence that an article investigating particular framings of poverty in education policy neighbors an article focused on state-funded, school-community partnerships. As public institutions, schools are often the site of policymaking for members of the community. They are where policies discussed, debated, created, and disseminated meet parents and children in their everyday lives. For educators, schools are the sites of an often personal, yet professional kind of politics, as language is used to construct and resist particular identities, subjectivities, and rationalities for students as learners and as political beings. Within a Bakhtinian analysis, informed by CDA, Wilinski demonstrates the fraught realities of hyper-rational ideals for school-community partnerships, which too often whitewash and skim over the differences in power, position, and discourse within and between partnered organizations. It is within these differences in discourse that she finds insight about implementation and partnership-in-action. Rather than pronouncing a partnership successful or failed because of its outcomes, she identifies and investigates the internally persuasive discourses participants shape in response to the authoritative discourses of policy. In doing so, she highlights findings about the nature and possibilities of legislated partnerships, as well as the meanings produced by participants over time.
In the third article, Koyoma returns to students in U.S .public schools later in their school careers when they transition from being the object of policy to active participants in the policymaking process. This article takes on questions about the nature of civic education and civic identify formation, specifically voting participation among Latino youth. Using CDA and Actor-Network Theory in combination, she examines civics education policy in juxtaposition with the enactment of citizenship by Latino youths. As in Wilinski's article, this juxtaposition of the authoritative discourses of policy with the assembled discourses of students who are the objects of that policy reveals how language and action take up, reframe, and resist policies at the site of implementation. This examination does not end at description, but rather leads to a discussion of implications for what policy could be, using the discursive enactments of citizenship to imagine a different sociopolitical reality as policy. In this way, the article not only contributes to our understanding of discourse and its function within policymaking and implementation, but also points to the potential genesis of policy discourses.
As Koyama outlines a possible future for policy, constructed from the citizenship enactments of historically disenfranchised youth, in the fourth article, Hurst examines another method of ongoing policy construction. Hurst's analysis of district superintendents' political tweets marks a departure from analyses at the nexus of authoritative discourse of state policies enacted by local stakeholders or citizens. Twitter data certainly brings with it radically different conventions than traditional state policy writing, particularly as the temporal relationships between traditional policy statements delivered by press release vary from the short tweets common to Twitter that come in sequence, punctuating events over time, with some room for interaction via comments, retweets, and likes. Hurst argues that it is within this mode of communication that district superintendents find ways to connect directly with stakeholders and therefore to communicate subjectivities, positions, and policy. Drawing on discursive psychology, Hurst identifies how superintendents use language to represent their engagement with the public, presenting themselves as engaged with constituents and colleagues, in the policymaking process, and even in activism. He argues that this presentation of self via Twitter allows superintendents to show what they do and what they value in real time rather than only in public meetings or on a campaign trail. Hurst notes that Twitter is becoming a platform for political identity and will therefore continue to be an important site for inquiry related to political engagement and the superintendency.
Indeed, Twitter is a site whereby personal political platforms are constructed, but it also a site for the development, dissemination, and debate of policy issues. In the fifth article, Supovitz and Reinkordt draw on a corpus of tweets related to the Common Core State Standards (CCSS) in order analyze the frames, metaphors, and lexical choices that are used to present the CCSS as a policy that required broad public opposition. The analysis of how language is used to frame CCSS as a policy issue demonstrates how particular frames might appeal to the value systems of both conservatives and liberals, thus creating a broad coalition to oppose a policy that might otherwise have been understood as a partisan issue. Unlike the other articles in this special issue, this article begins and ends within public discourse about policy, examining how it shapes and limits possibilities for implementation. It therefore occupies the in-between space between policy formation and implementation wherein public opinion can work to shift the very nature and direction of policy. In this sense, it is not only the use of language-based data drawn from tweets that place this work squarely in the 21st century, but also the interest in the power of public response on social media as politics and policies are increasingly communicated and contested virtually. Supovitz and Reinkordt's study serves to underscore Koyama's suggestion that current civic engagement curricula sharply underestimates the role and importance of electronic and virtual participation, and that a revised curriculum for civic engagement should examine the many ways of being politically active online.
The final article in this special issue examines a focused intersection of policies: where higher education policies meet immigration policies. Gildersleeve draws on data from popular journalism and presidential campaigns to describe the subject positions made available for understanding students as having "undocumented" and/or "illegal" identities. This analysis eschews the notion of a single or fixed identity and instead examines how possible selves are made plausible and relevant as various subject positions are produced in and through discourse. Drawing upon post-qualitative understandings of research, Gildersleeve highlights the ways in which people come to be understood and positioned as "undocumented" or "illegal" are coordinated discursive accomplishments, not biological or non-negotiable realities. He highlights how it is within these discursive productions of possible selves that political power over individuals and their lives is asserted. This understanding of subject positions instead of or beyond the humanistic notion of identity is undertaken within a theoretical approach named post-qualitative, and a methodological approach named specifically as "policy discourse analysis," within which policy is always already understood as discourse which produces particular truths and understandings that are never as stable, unified or self-evident as they may seem.

Policy Implications, Future Directions, & Conclusion
Collectively, the articles included within this special issue highlight the various ways in which policy issues, problems, and responses reside at the intersection of political and organizational identities. Discursive perspectives thus provide a valuable means of analyzing these identities and unpacking how they influence both the policy formation and implementation process. Within these articles we see how discourse is used to 'mark' and 'de-mark' particular constituencies, interest groups, and stakeholders. Though policy scholars have investigated how particular policy issues are identified within the policy process, we see within these articles how important language is within this process. This has clear implications for the study of policy as it is ultimately in the process of identifying and denoting stakeholders that we, as policy scholars, come to know why policies exist as they do and what their existence means for the stakeholders they impact. By way of recent example, the stunning shift in education policy discourse as it relates to choice and vouchers between the Obama and Trump administration highlights just how rapidly changes in language signal potential shifts in policy.
The included articles also illuminate the changing nature of policy-relevant discourses and the sources of evidence that can be used to probe and problematize these discourses. For instance, both Hurst's and Supovitz and Reinkordt's articles signal the rapidly rising importance of alternative forms of policy communication. Twitter, for example, has increasingly been recognized as a form of policy discourse that not only can signal to relatively uninformed stakeholders misguided interpretations of policy but also may be used to stimulate fear, apathy, and activism among policyinterested stakeholders. While policy scholars have increasingly been studying the use of Twitter, the articles in this special issue highlight how this might be undertaken using discursive approaches. Indeed, we think one of the key contributions to policy research that these articles might have is by way of modeling how policy scholars might tap into new and emergent data sources that have historically been subjugated to a lesser status than more traditional forms of qualitative data, such as interviews, observations, and documents. Expanding the pool of potential data sources in policy research is thus one of the chief contributions that this special issue makes. Indeed, we hope that other scholars will examine how policy and other education stakeholders use social media to influence the debate about what happens in schools. This has direct bearing on the policy process as it is often through policy-makers' observations about these practices that new policies are formed.
On methodological grounds, these articles push the boundaries of policy research beyond its dominant use of CDA when studying language. As we noted above, CDA has been the dominant approach used within education policy. Perhaps this is unsurprising in that CDA has historically foregrounded the importance of attending to power, (in)equality, and dominance within political realms, as produced in text and talk (van Dijk, 1993). Indeed, we recognize that CDA itself represents a diverse set of approaches to studying language, including critical linguistics, sociocognitive perspectives (van Dijk, 2001), a discourse-historical approach (Wodak, 2001), Fairclough's approach (Fairclough, 2013), among others (Faiclough, Mulderrig, & Wodak, 2011). Like Taylor (2004), we suggest that CDA approaches are particularly useful for critical policy research. Yet, as this special issue highlights, there are many other discourse analytic perspectives that offer unique pathways for studying language at both a macro and micro level. These diverse perspectives include differing conceptualizations of what counts as language, how discourse should be defined, and even how and where an analyst should focus their analysis. For instance, within discursive psychology (Edwards & Potter, 1992), it is not uncommon for researchers to attend to both what is said and how it is said, pursuing what some might call a more 'micro' level of analysis. In other words, in some micro-oriented discourse analytic approaches, paralinguistic features of talk, such as pauses/gaps, intonation, etc., may be assumed to be important and therefore attended to within the analytical process. As another example, a Foucauldian-informed discourse analysis (Arribas-Ayloon & Walkerdine, 2008) brings to the fore particular analytic foci, including a focus on historical or genealogical analyses, power, and subjectification. Certainly the interdisciplinary landscape of discourse analysis includes overlap across perspectives; however, the uniqueness of the various discourse analytic approaches afford policy scholars opportunities to work in differing ways at the intersection of education policy and discourse analysis. Thus, we see this special issue as extending the methodological boundaries for how policy scholars might go about studying the ways in which policy is constructed, appropriated, and implemented.
Much as we did in the first part of this special issue, we conclude by calling upon policy scholars to continue exploring the applicability of discourse perspectives to the study of education policy. We acknowledge that methodology does not stand still and that the advances presented in this special issue may well be eclipsed as further innovations occur. Thus, we hope that the articles within this special issue encourage policy scholars to consider other boundaries that may be crossed methodologically and substantively, as such crossings may very well advance the needs of the field of education policy and lead to new understandings and practices.