An Anthropological Approach to Education Policy as a Practice of Power: Concepts and Methods

Since the introduction to our 2001 edited volume, Policy as Practice: Toward a Comparative Sociocultural Analysis of Education Policy (Sutton and Levinson 2001), we have continued to sketch the foundational postulates of a critical anthropological approach to the study of education policy. In 2009, we expanded and deepened many of the points from that introduction, more systematically introducing and defining theoretical terms, and providing a bit of their intellectual genealogy (Levinson et al. 2009). We also discussed certain methodological considerations that accompanied the theoretical approach, and we argued for a type of engaged educational anthropology that goes beyond the mere “study” of education policy to its democratization and transformation. Here we provide an updated synopsis of our approach.

produced democratically-as much for various civic publics as for scholarly communities, in addition to those we call "authorized" policymakers.

Anthropological Foundations
Arguably more than any other field, anthropology demands that we understand any contemporary social phenomenon in the context of history and the illuminating prism of cross-cultural diversity. What this means practically for our case here is that we must constantly examine and decenter our common assumptions about what "policy" is or does. Policy has only entered the English vernacular within the last 100 years or so, yet it is now invoked across a wide array of social and political contexts, and it has developed numerous variants and cognates in other world languages. So the first step is to realize that policy as both an English word and (post) modern tool of governance has a particular history and a particular cultural location. We can trace its origins most broadly to the Enlightenment's emphasis on rational social engineering, and over the course of the twentieth century to an increasingly technocratic rendering of liberal democratic governance; we can also trace its spread more recently through other languages and societies, both of the ethno-national and popular kind, where it often acquires new usages and inflections. Indeed, it was only when Levinson's young daughters used the term "policy" to describe the rules they had created for an invented walking game, "Don't Step on the Sidewalk Cracks," that the full scope of its use became apparent.
So what is this thing we call policy, which gets promulgated and negotiated amongst supranational organizations, nation-state provinces, grade-school playgroups, and seemingly everything in between? What are its common traits? We think it best to define policy as a complex, ongoing social practice of normative discursive cultural production constituted by diverse actors across diverse contexts. The resulting normative cultural discourse has positive and negative sanctions, that is, a set of organizing principles about how things should or must be done, with corresponding inducements or punishments. Such a discourse may or may not be formalized and codified (it could be stated or unstated, explicit or implicit, de jure or de facto). In every case, though, it crucially presupposes a view of how things "are"-a model of the world that is being organized, as it were-and how they "should be." In order to solve practical and existential problems, policy thus defines reality, organizes behavior, and allocates resources accordingly.
The anthropological lens insists on policy being a kind of link between the discursive practices comprising larger-scale structures of law and governance and the discursive practices of normative organizing and control in any local-level site or community. Policy, then, lives in the liminal space between cultural norms and laws; it is more actively codified perhaps than norms, less binding and coercive than laws. Importantly, policy organizes social settings whose actors may have quite different levels of awareness or agreement about how they are being organized. The adult who makes a living teaching second grade and the seven-year-old in her second grade class may not understand the world all that similarly, nor agree on what is supposed to happen in the social environment they share, but they are both nonetheless organized by policy.
Important work in the fields of political and legal anthropology can clarify our approach here. Indeed, what we've just described as the liminal space where norms are transformed or codified has been the focus of the field of legal anthropology, and its evolving concept of legal pluralism. The concept of legal pluralism takes the connections between multiple legal systems seriously by focusing on the dialectical relationship between state systems of legal order and other normative social orders, usually at the tribal or community level. Sally Falk Moore (1973) describes how laws or policies operate in a semi-autonomous social field which …can generate rules and customs internally, but that…is also vulnerable to rules and decisions and other forces emanating from the larger world by which it is surrounded. The semi-autonomous social field has rule-making capacities, and the means to induce or coerce compliance; but it is simultaneously set in a larger social matrix which can, and does, affect and invade it, sometimes at its own instance. (p. 720) Moore's work speaks to the multi-locality of the practice of policy, and echoes our emphasis on policy as a practice of power as well as a vehicle for potential resistance. Her work calls our attention to both the larger dynamics, created by state-enforceable law, that create and organize social life, and the smaller dynamics-the uncodified orders and scripts or the implicit rules for engagement in a particular social context-operating in between larger social forces. She reminds us that these ordering principles, large and small, are interconnected and cannot be separated in our efforts to understand how law and policy organize social life. Even when policy actors are appropriating policy, acting in informal spaces, and resisting the coercive power of more official forms of policy, they do so within the larger social matrix.

Practice, Practice, Practice
As we hope to have shown by now, anthropology helps to decenter the seeming naturalness and "normalcy" of policy in the current historical moment. It enables us to question our taken-for-granted sense of policy. We further suggest that the way to unpack policy analytically is to see it as a kind of social practice-specifically, a practice of wielding power in modern forms of governance. What might it mean to take policy as social practice? How can we put action and agency back into a seemingly static policy text and conceptualize the entire policy process as a complex set of interdependent sociocultural practices? How might we look beyond text to examine critically the ways in which official, top-down policy practices are negotiated, contested, accommodated, or transformed in action?
Even though policy most often takes the form of discourse, or text, we prefer to put this form into motion and always analyze it as one reified instance of a broad chain of sociocultural practices. On the one hand, policy as normative discourse may be what we call officially "authorized"-that is, backed by enforcement mechanisms of government, organizational, or corporate charter. On the other hand, policy may also develop in more spontaneous and informal fashion, outside the agencies or offices that are officially authorized to make explicit policy. Those subject to the vertical or diffuse power of authorized policy may well assert their own power in response. In either case, policy may be documented and codified, or it may exist in "unwritten" form, through ongoing institutional memory and practice.
In every instance, policy formation is best conceived as a practice of wielding power. Such a practice of power may be more or less democratic, depending on the ways that power elites are formed and legitimated, and the ways that publics are constituted to participate (or not) in policy formation. In its most common and visible forms, authorized policy under advanced capitalism is preeminently modern, characterized often by a rationalist calculus and a representative democratic veneer. Despite the rise of supranational organizations under globalization, the modern state apparatus, of course, is still the supreme authorizer of policy. And state "public policy," especially, manages to obfuscate its reality as a highly political form because it is effectively disguised by seemingly "objective, neutral, legal-rational idioms" (Shore and Wright 1997).
From the 1950s forward, traditional policy studies, influenced heavily by the political and economic sciences, have focused on the strategic conditions for policy formation and implementation, as well as methods for analyzing policy impact. We problematize this emphasis and foreground instead the social practice that goes into the formation, negotiation, and appropriation of policy. A processual practice approach takes less for granted about the presumed rationality of "problem identification" in the formation of authorized policy. It looks more closely, instead, at the social arenas where the interests and languages comprising a normative policy discourse get negotiated into some politically and culturally viable form.
The concept of negotiation, of course, has an overt political reference, and we mean to recognize that authorized policy does often get "negotiated" between opposing parties and interests. However we also wish to highlight the term's sociocultural sense, that is, as a way to account for the processes of meaning-making. In addition to possibly negotiating a "deal," authorized policymakers negotiate a complex field of meanings and understandings. According to this anthropological perspective, the making of meaning is fundamental to social action, and meaning is therefore always "negotiated" in social life; values are never fixed but rather are contingent on the mobilization of meaning in specific situations. For us, the negotiation of meaning is always a part of policy formation, whether or not actual political negotiation was involved; in other words, the process of normative cultural production requires an active negotiation of meaning.
Beyond the process of policy formation itself, the negotiation of meaning occurs across and within the various institutional and organizational sites where policy flows and takes shape. Instead of using the term implementation to analyze this process, we propose the concept of appropriation, which places emphasis on how social actors take something initially external to their social context and make sense of it. Appropriation is a concept that has a complex history. It has been developed and used in the work of Marxist phenomenologist Agnes Heller, cultural studies of the media, Bakhtinian discourse analysis, legal scholarship on intellectual property rights, and so forth. Across these various traditions, appropriation denotes how the individual or the group reshapes and resignifies in practice some previously existing cultural artifact. Appropriation thus refers to the way that creative agents interpret and "take in" elements of policy, thereby incorporating these discursive resources into their own schemes of interest, motivation, and action, their own "figured worlds" (Holland et al. 1998).
The study of policy appropriation has tended to highlight later moments of the policy process, when the authorized text or "policy signal" circulates, by various means, across the various institutional contexts to which it applies. Admittedly, our earlier work has been read as suggesting that there is an originating policy formation process, with appropriation only taking place during later stages of an implementation cycle (Nielsen 2011). However, we wish to emphasize that appropriation, like negotiation, takes place during processes of authorized policy formation as well. We insist that authorized policymakers are appropriating discourses and narratives through their own practice of power. Appropriation thus should not only be used to characterize the actions of the so-called implementers of a policy that has been formed officially. Rather, appropriation is always part of the practice of power in authorized policy formation, too. It may be that authorized policymakers are freer to selectively adapt, borrow, invent, dissimulate, or otherwise manipulate their policy ideas, but they are appropriating them from somewhere, not nowhere. 1 As we say, it's practice all the way down, up, and through the social world. It's also resignification all the way down, up, and through.
The concept of appropriation is vitally linked with a broader conception of social and cultural practice that in recent years has come to invigorate the human sciences. We see two primary streams of work comprising such "practice theory." On the one hand, sociological and anthropological theorists such as Giddens (1979Giddens ( , 1984Giddens ( , 1991, Bourdieu (1977Bourdieu ( , 1990a, Connell (1983Connell ( , 1987, and Ortner (2006) have developed conceptions of practice to resolve perennial antinomies between structure and agency, or society and the individual. In their accounts, social practice is the "site" or "moment" where structure and agency, individual and society, mutually constitute one another. Meanwhile, out of psychology, work on "situated cognition" and "activity theory" has tried to resolve similar antinomies between "mind" and society, or cognition and environment. Influenced by the Russian sociohistorical tradition, especially represented by Lev Vygotsky, theorists such as James Wertsch (1991), Michael Cole (1996, and Jean Lave and Etienne Wenger (1991) have articulated powerful new models of human thinking and learning that account fully for the inherent and emergent social properties of mind. Together these approaches serve to remind us that education policy is simultaneously "inside and outside" of the educators and students who enact it. As participants in an educational system, we can often more easily identify the external signs of policy than those deeply internalized, so our work may need to tack back and forth between visible policy signals and the way such signals get organized as implicit cultural knowledge and then manifest in behavior. Moreover, structures of power, like policy, are not merely imposed from without but often subtly instantiated in everyday belief and practice (Levinson et al. 2011).
By the same token, we cannot say in each instance that policy appropriation by non-authorized actors recursively links back to, or influences, authorized policy formation. Some critics of our work have said that our emphasis on the recursive nature of social practice, and how unauthorized actors form their own localized policies in their own spheres of action, runs the risk of "flattening" our understanding of the full policy process and overstating the power of unauthorized actors to "make policy." To be sure, we need to keep in mind the distinction between official, authorized policy formation (conducted by elected and appointed officials, bureaucrats, etc.) and the unofficial practices of unauthorized actors, like teachers. We can't remain innocent of the very real power that authorized policymakers have, and we can't wax too optimistic about the ability of non-authorized actors to influence consequentially the ongoing formation of official policy. These are empirical questions and matters, to be investigated through research.
Indeed, recent work in the field shows quite trenchantly how even the notion of who is authorized, or in what authorization consists, is often deeply contested (Lashaw 2018;Sandler 2018). The social dynamics of policy authorization cannot be taken for granted, or simply ascribed to dominant state and corporate actors. Moreover, Bruno Latour's Actor Network Theory draws our attention increasingly to the role of nonhuman and even nonliving "actors" in the "emergence" of policy (Koyama 2018). Until now, our analytic move has been to theorize the agency of non-authorized actors, to recognize where such agency has progressive consequences, and to explore the conditions that enable interaction between authorized and non-authorized policymakers such that authorized policy that has been illconceived or undemocratically imposed can be effectively contested or changed. Yet this more recent work challenges us to keep expanding our understanding of agency, and to stay attuned to the complications of power, especially in collaborative, action-oriented policy studies.

A Brief Example
Winstead's recent work, on Indigenous education policy in the state of Washington, U.S.A. (Winstead 2014), provides an example of the multiple meanings differently positioned policy actors may make in the policy process. In this example, some of the state policy makers understood the policy initiative-to include Indigenous history in high school graduate requirements (HB 1495) 2 -as based on recognition of the inherent sovereignty and the long political history of the tribes in this region. Another group of policy actors negotiated and interpreted the meaning of the same policy in a different way. Utilizing a liberal multicultural lens, these policy makers understood the objective of the policy as an imperative to incorporate Indigenous history into "the story" of the state's history. They were advised to delay action on it because legal advisors saw it as a potential precedent for other minority groups' curriculum to be included in the official policy lexicon. Analysis of the transformation of this bill as it passed through the legislative process, coupled with ethnographic interviews with the policy actors who held competing visions of the purposes of the bill, revealed the complexity of the intersection of state education policy, Indigenous and non-Indigenous understandings of sovereignty, and broader efforts to decolonize American Indian education. We think this example illustrates how the policy process plays out as a complex set of interdependent sociocultural practices informed by differing sociopolitical scripts for meaning-making and cultural appropriation.
In addition, thinking in this way, like the legal pluralists who inform our approach (Merry 1992;Moore 1973;de Sousa Santos 1987), focuses our attention on how unequal but mutually constitutive legal orders interact. This approach is particularly well suited to the study of the relationship between federal and state legal processes and those of subordinate groups: in this case, Indigenous advocates for the inclusion of tribal history in public school curriculum. Situating the proposal made by HB 1495 within this nexus of structural arrangements, it becomes clearer that the initial failure of the bill (in its first form as HB 2406) and the eventual amendment of the bill to remove the mandatory teaching of native history in Washington, can be understood as a process influenced by hegemonic forces of ideological reproduction. That is, teaching native history (as HB 1495 promotes) is counter-hegemonic, and using the lens of practice theory, in the way we suggest here, helps to identify and explain why this legislation evolved in the way it did.
Santos identifies the ways in which law operates like a system of signs that represent and distort reality "through the mechanisms of scale, projection, and symbolization" (de Sousa Santos 1987: 297). In following Santos' argument, Merry (1992) points out that this approach prompts research which asks questions about the degree to which the dominant system is able to control the subordinate, and to what extent and how subordinate systems evade, resist, or "invade" the dominant system.
When two legal world views interact, radically different views of what constitutes proper government-to-government relations-especially between Native and non-Native individuals-can result. This makes communication about policy intentions fraught with potential misunderstanding, which may be due to differences that are related both to ideological and culturally constituted commitments. A policy as practice approach considers these related factors about how policy actors involved in the passage and implementation of H.B. 1495 fit within the ideological and hegemonic structure of the social world, and how that impacts the potential for success of the policy in terms of its goals to foster student success, educate the public, and facilitate government to government relations.
Much of the logic from the state policy apparatus focused on the need to maintain local control over curriculum at the district and school level, and the risk to the integrity of the policy lexicon in terms of representation of one minority group in the state's required curricular content. The rationale for changing the policy from a mandate to an "encouragement," in public testimony and interviews with policy actors, was that the legislature did not possess the authority to change HS graduation requirements, because Washington is a local control state. Policy makers at the State Board of Education were concerned that granting this request would "pave the way" for additional requests of this kind from every minority group in the state. That is, they viewed the tribal reform initiative primarily through a liberal multicultural framework, and secondarily through a framework recognizing the political sovereignty of tribal nations (see Table 17.1). Senator McAuliffe, the chair of the Senate Education Committee, responded to this testimony by explaining why the language was changed from mandate to encouragement. This explanation is the only one that is offered throughout the discussion of the bill: The reason we changed Section 3 and have the State Board of Education considering tribal history and culture in the graduation requirements is because the State Board of Education is responsible for graduation requirements, so rather than have the legislature dictate what would be required it is the State Board of Education's responsibility, so we have given it to them to make due requirements… and that's their job…So, it really didn't weaken it, it just sent it to the right body. [Public Testimony, Senate Education Committee, March 25, 2005] This moment in the process of the evolution of the legislation is pivotal because McAuliffe's perspective points to the official policy reasoning behind the movement from mandate to encouragement; it also points to the bureaucratic vision that this legislative body has about how legislative authority should be used in this case. Senator McAuliffe, here, indicates that this is really the best and only option that they, the authorized policy makers have, to achieve the intentions of the bill, because this falls under the authority of the State Board of Education.
The contrasting framework was promoted by advocates for HB 1495, and emphasized, among other things, the inherent sovereignty of the tribes, and the need for relationship building between tribes and state agencies and school districts (see Table 17.1). This perspective is illustrated in the following quote from Representative McCoy, the bill's sponsor, who introduced the policy's intention to the House Education Committee during a pre-legislative working group: McCoy's summary is a direct address to non-Native policy makers, and points to the friction between the process-oriented request for the inclusion of indigenous history and culture into the official policy framework, and a more robust political request, for tribes to fully participate in the policy process.
The way indigenous history was initially handled as a possible state-sanctioned addition to Washington curriculum was not arbitrary. It is directly related to the presumptions of the state, and the room allowed for indigenous content in official state policy spaces. This is the logic written by the settler colonial history of this country, 3 which underwrites the narratives of American education policies. These narratives delimit the normative range within which conversations about indigenous education can occur. They do not, as such, operate as fully written dominant narratives, but there is a sense in which the Bordieuan doxa (1990a, 1990b)-that which is taken for granted-of the state policy-making field does not admit indigenous agendas into its lexicon or narrative record easily. Each phase of the debate about the inclusion of indigenous education in Washington common schools leaves its mark and contributes to the next phase of coordinated attempts to challenge the state's exclusion of indigenous history into the graduation requirements in Washington State's common schools.

Beyond Provocation: Some Final Notes on Methodology and a Democratic Ethics
Until now, much of what we have argued would hold true across many different types and domains of policy: education, but also health, economic, environmental, and so forth. But there is something about education that makes it singularly fraught as a field of human endeavor. Education is universally practiced and deeply intimate. It encompasses the production and transmission of cultural knowledge across generations, and it serves as the crucible by which human beings learn to make their way in the world, first and foremost in the company of kin. Everyone educates, but only some become professional educators, and therein lays the rub. The professionalization of education and the historical rise of schooling as a political tool of the state bring large-scale structural imperatives into the bosom of the family, so to speak. Official state education policy attempts to reach into and organize the family/ community and its enculturation of the self. It often does so in pursuit of national "unity" or "security" or "development," not to mention the specific interests of dominant groups in reproducing the status quo. To speak, then, of democratizing education policy is to envision the ways that ordinary people across a society's structured inequalities can gain a greater measure of control over their children's education and challenge insidious forms of assimilation or social reproduction. Yet even as we offer such a formulation, we must be wary of our penchant for romanticized localism. Just as the state does not always reproduce inequality, local groups do not always liberate themselves or a broader public. Parents and communities, alas, do not always act in their own children's best interests (Lashaw 2018). An obstinate emphasis on local control or the superiority of local cultural knowledge can lead to the dismissal of valuable, even crucial forms of knowledge offered by schools. There is, after all, good reason for the professionalization of schoolteachers. Thus, education policy democratization goes both ways. Yes, it stands for the promise of empowering parents and other local actors-including teachers themselves, who are rarely consulted-to have a much greater voice in state-level policy formation. But it also means that professional educators must be willing to dialogue with children and parents and community leaders, to temper their assertion of professional privilege with a dose of humility. And it also means that students and their parents must commit to the public good and measure their own particular interests against that yardstick.
We would like to use the distinguished anthropologist Michael Herzfeld as a bit of a foil here to argue for an anthropology of education policy that regularly goes beyond critique and provocation. In the preface to his 2001 book, Anthropology: Theoretical Practice in Culture and Society, Herzfeld says that he offers a model for a "critical engagement with the world" (p. x). Such engagement, he says, may take many forms, from critique of policy or professional practice, to "ethnographically oriented phenomenology" (p. x). Most tellingly, Herzfeld argues that anthropology is a "provocation, not a prescription: that is the kind of teaching that anthropology offers, which is why it is so disapproved by normativists of all stripes-official ideologues, econometric modelers, champions of western (or any other) cultural dominance" (p. xi). To be sure, anthropology can and should play a powerful role in social critique and "provocation," but we don't wish to leave it at that. After all, not all "normativists" can be so neatly painted in diabolical terms-some of them are democratic socialists in municipal government, or progressive school reformers and school superintendents, or teacher educators. Some of them, by golly, are us!! So the question that emerges is not whether the anthropology of education should "prescribe" policy changes based on its research; after all, such a move would be presumptuous, and violates our understanding of the importance of context. But in between provocation and prescription, there is the option to provide knowledge that is normatively digestible. In other words, the anthropology of education can offer knowledge that is politically and administratively feasible and actionable. If not, there is something all too comfortable about occupying the space of critique in the academy; we either tend to exaggerate the potential power and influence of academic critique, or we wallow, frankly, in a kind of smug self-righteousness about how we can really see what's going on behind people's backs. Through the way we write and speak, we may reproduce the self-fulfilling prophecy of our own irrelevance to broader worlds of policy and practice, and come to feel comfortable and even superior in that irrelevance.
We also mustn't forget that there's a whole world of practicing educational anthropologists beyond the academy, and they don't have the luxury of producing knowledge only for critique (Schensul 2011). But for those of us who are located in the academy: What happens when we venture into the messy world of policy, practice, and governance, where we can't just critique but must propose, if not prescribe? What happens when we align ourselves with social movements in outright advocacy or activism? What risks can we take, and what risks should we take? More to the point, how can a new, fundamentally ethnographic approach to research on "policy as a practice of power" contribute to democratizing policy formation, in which policy elites must necessarily understand local knowledge and engage local stakeholders, and in which local policy production can be catalyzed? In sum, if policy is a practice of power-that is, of defining reality and organizing behaviorthen how can ethnography be a practice of questioning dominant definitions, proffering alternative ones, and reorganizing (transforming?) behavior and society?
The path toward answering these questions, we argue, must be walked with an eclectic, pragmatic vision, informed by values of inclusivity and respect. Such a vision expands the purview of participation and justice in a liberal representative democracy, by making it more possible for a greater variety of voices and perspectives to take part in deliberation, and to be represented in the formation of authorized policy. Yet we also recognize and critique the limitations of representative democracy and see ethnography as contributing to the empowerment of historically subordinated people, helping to expand the reach of participatory democracy (Sader 2005) by catalyzing popular agency and knowledge for creative policy appropriation and production. In particular, the anthropology of education policy enables us to insert knowledge into different circuits of appropriation and social mobilization. At a time when democracy is threatened worldwide by a neoliberal agenda of corporate power (Graeber 2009;Kirsch 2014) and the marketization of public resources (Brown 2005;Giroux 2015), critical policy appropriation can bolster the agency of local groups and institutions struggling to restore or expand public power.
In our own field of educational anthropology, there are far too many promising examples of such work to adequately cite. The pioneering work done by Norma González and her collaborators on "funds of knowledge" (Gonzalez et al. 2011) has acted to democratize the prospects for Mexican American student empowerment in Arizona and elsewhere. The deficit perspective informing most school curriculum policy is challenged by the demonstration of rich student and familial funds of knowledge. Angela Valenzuela's work in the Texas legislature (Lopez et al. 2011) shows how we can intervene with specific ethnographic knowledge to alter important policy debates and reconfigure the calculus of educational resource distribution; it also shows how we can undertake action-oriented ethnography to better understand and change the limits of democratic representation in such policymaking bodies. In a different sphere, pioneers in participatory action research with youth (Ginwright et al. 2006) show how policy can be questioned, destabilized, and in some cases changed democratically through sustained and passionate inquiry. Finally, in Levinson's work on Mexican secondary reform, his talks and writings across various educational constituencies in that country have attempted to foster greater input and consideration of teachers', parents', and students' voices in the policy process (Levinson 2007(Levinson , 2008Levinson et al. 2013). Levinson thereby strives to insert his research findings into the struggle between a neoliberal project of so-called accountability and standardization and a long-lived popular project for democratization from below.
We wish to expand the space of the public-and speak to that public, not just to policy elites. Like some authors (Walters 2000), we believe it is crucial to question the privileged status of scientific or expert views and to reinvigorate public involvement in the policy process. The work of critical ethnographers (Carspecken 1995;Madison 2011) is similarly conscious of power and democracy, and in effect, we also wish to redefine the goals of interpretive research-away from a strictly academic practice of theory development, or an academic practice of "influencing" authorized policy, and toward engagement with the aims of democratic social movements (Appadurai 2000). Such scholarship for popular democracy may well have as one of its goals the development of a cultural critique, theorizing the way that "controlling processes" (Nader 1995) limit and blunt the full possibilities for democratic participation. Yet an anthropology of education policy may also contribute knowledge to alternative democratic projects, to educational efforts aimed at creating plural "counterpublics" for a democratic renaissance (Benhabib 1996;Fraser 1989).
We also relate our approach to the earlier discourse, originally introduced by Laura Nader (1969) about studying "up" the power structure versus studying down. A policy as practice approach takes seriously the need for critical work that "studies up" and uncovers the strategies and mechanisms at work in elite, authorized policy formation processes. Such knowledge can then be circulated to democratic actors situated in other social domains, and in social movements, to foster greater accountability and strategic mobilization. Yet equally important is work that "studies down" to understand how marginalized, powerless groups, not authorized to make official policy, nonetheless create policy variants through their appropriation of authorized policy. Finally, we must study "through" and "across" both horizontally and institutionally linked groups and organizations, such as the intermediaries studied by Ted Hamann and Brett Lane (2004).
Lesley Bartlett and Frances Vavrus' (2017) brilliant approach to comparative case study, which urges attention to the "horizontal," "vertical," and "transversal" dimensions of education policy processes, provides a comprehensive methodological and meta-theoretical companion to our work here. Like them, we argue for a continual redefinition of the ethnographic "field" away from the earlier conceits of "bounded" cases and places. Because policy is fundamentally a normative discourse, and often intends to order or control relations between groups that may occupy very different social spaces and scales, research must be "multi-sited" and "multi-scalar," attuned to the production, flow, and appropriation of reified texts across time and place. The new "field" of qualitative research is thus neither the traditional community nor institution, but rather a constellation of social sites, contexts, and networks. To be sure, many of these social sites may be parallel in status (e.g., a set of similar schools that are all subject to the same policy), and thus lend themselves to what Bartlett and Vavrus call horizontal comparison, or what we have called a comparative latitudinal approach that enables a kind of sampling of forms of policy appropriation (for instance, Levinson's strategy in his recent work in Mexico (Levinson et al. 2013), or Winstead's (2014) work tracking policy artifacts through state legislative processes in Washington State).
Similarly, "vertical" analysis calls attention to policy processes that cross scales of power and governance. There are many techniques for conducting qualitative research across scales: these include observational and interview work in a range of interconnected social sites across such scales (e.g., international organizations, national education ministries, regional education authorities, and individual schools or classrooms). 4 At the very least, vertical analysis calls for attunement to the structuring power of policy processes that are scaled up and out, even when we are engaged primarily in rather local data collection. Correspondingly, we urge new research practices of institutional and discursive mapping, in which policy language is traced across documents (Lester et al. 2017), and in which graphic representations are made of the quality and density of actual relations between institutions and actors that produce policy, and those that appropriate it. 5 Finally, we believe that now more than ever, attention to the historical structuring of policy processes (what Bartlett and Vavrus call "transversal" analysis) is critical. This can range from the long dureé to a more compact (3-20 years) unfolding of policy. For the latter, longitudinal research designs are necessary to capture the fullness of the policy process. When possible, the researcher ought to be present at the early stages of policy formation to observe the relations and interests that emerge; if this is not possible, the practice of policy formation can be reconstructed through qualitative interviewing. Then, depending on the policy in question, the research design should include at least a 3-4 years investigation of the "life" of the policy as it gets implemented across various sites. Such a minimal time frame would enable better understanding of the processual aspects, the unanticipated twists and turns, of policy appropriation; it would bring into view the possible recursive aspects of policy appropriation, which may eventuate in modifications to the authorized policies themselves (i.e., as a form of "policy learning"); and it would foster the forging of deep relationships between the researcher and the subjects of policy, to facilitate the researcher's role as mediator and translator in policy democratization.
Taken as a whole, such a research agenda can move us beyond provocation and critique. Methodological innovation, long-term commitment, and knowledge for advocacy and democratic policy (re)formation: these are crucial ingredients in any recipe for progressive education reform.
Bradley A. Levinson is a Professor of Education in the Department of Educational Leadership and Policy Studies and an Adjunct Professor of Anthropology and Latino Studies at Indiana University. A Latin Americanist by training, his empirical research focuses on secondary-level student culture and identity formation, democratic citizenship education, and the politics and culture of education reform. Recent projects include Mexican secondary education reform, models for radical civic-ecological identity formation, and Spanish-English translation of Latin American educational research representing alternative epistemologies. Teresa Winstead is an Assistant Professor of Anthropology and Sociology in the Department of Society and Social Justice at Saint Martin's University in Lacey, Washington. Winstead is an anthropologist of education whose research focuses on how policy impacts educational practice, from teacher perspectives to student experiences. Her work has dealt primarily with the context of Indigenous education policy in Washington State. Over the last 3 years, Winstead has partnered with Washington State's Office of Native Education on several research projects, investigating how policy actors at various levels of the policy process interpret and implement existing policy intended to foster increased engagement and student success among native youth in Washington's schools. Currently, Winstead is working with the Office of Native Education and The Evergreen State College's Enduring Legacies Native Case Studies Institute on a collaborative project which will document the implementation of the Tribal Compact School legislation.

Margaret Sutton is Associate Professor of Educational Leadership and Policy Studies and an
Adjunct Professor of Global and International Studies at Indiana University. An anthropologist and comparative educator, Sutton has worked extensively in the field of education and development, and has written about social scientific knowledge in post-colonial settings, gender and education, and curricular practices in international and global education. Her current work focuses on teacher education and higher education development in Indonesia.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.