Summary of "Frame Reflection: Toward the Resolution of Intractable Policy Controversies"

Summary of

Frame Reflection: Toward the Resolution of Intractable Policy Controversies

By Donald A. Schon and Martin Rein

Summary written by Conflict Research Consortium Staff

Citation: Frame Reflection: Toward the Resolution of Intractable Policy Controversies, Donald A. Schon and Martin Rein, (New York: Basic Books, 1994).

The authors argue that, contrary to much of the current opinion in the field, there is an important place for theoretical reflection within policy practice. Theorists such as Hanna Arendt, Joseph Gusfield, Albert Hirschman, Renata Mayntz, Jon Elster, and Thorstein Veblen have defended a sharp distinction between practice and reflection, and argued that "higher-level reflection in policy-making practice is neither feasible nor desirable."(xiii) Generally, such theorists argue that reflection requires distancing and disengaging oneself from the objects of thought. Reflecting on and questioning basic assumptions and values results in uncertainty, and so impairs the ability to act. In policy practice one must act decisively, and in order to act one must rely on some enabling assumptions. Reflection, then, is the antithesis of practice. Reflection should be left to the scholars.

This dichotomy between reflection and practice produces a dilemma. "Reflection of a kind that might hold potential for help in the resolution of intractable policy controversies is deemed to be out of place in policy making, where it might be most fruitful, while in the academy, which is seen as its proper locus, it tends to unfold in a way that is useless to those who are engaged in policy practice. On both counts practice loses out."(xvii)

Schon and Reid seek to develop an account of reflection in practice which will resolve the traditional split between reflection and practice. They are motivated to develop such an account by the problems presented by intractable policy controversies. Their text then is an attempt to respond to a basic set of questions: "How should we make sense of intractable policy controversies? How should we understand policy-making processes in which such controversies arise and persist? How should we account for the processes by which they are, or might be, resolved through reasoned discourse and reflection in policy making?"(22)

Part I: Setting the Problem of Reframing

Chapter One: Intractable Policy Controversies

The authors begin by distinguishing between policy disagreements and policy controversies.

  • Policy disagreements: disputes in which the parties to contention are able to resolve the questions at the heart of their disputes by examining the facts of the situation. For example, parties may disagree as to how many clients used a particular service. (3)
  • Policy controversies: disputes which are immune to resolution by appeal to facts. Such disputes tend to be intractable. For example, parties may disagree as to whether abortion should be legally available.(4)

Generally, we can distinguish between disagreements and controversies by the role that facts play in the dispute.(4-5) First, parties to a controversy often focus on different facts as being most relevant. For example, in the current controversy over abortion some people emphasize the potential human life of the fertilized egg, while others emphasis the right to self-determination of the woman. Second, parties to controversies tend to interpret the same facts in different ways. A decrease in the number of abortions performed may be seen as reflecting a decreased in the desire for the procedure among women, or as the result of legislative or practical barriers which make abortions increasingly difficult to obtain. The authors observe that "by focusing our attention on different facts and by interpreting the same facts in different ways, we have a remarkable ability, when we are embroiled in a controversy, to dismiss the evidence adduced by our antagonists."(5) Schon and Reid illustrate this phenomena by examine controversies over the sources of poverty and crime.(5-8)

Intractable policy controversies are damaging to the public in at least two ways. First, such controversies impair public learning because, the authors argue, " any attempt to conduct public inquiry into policy issues requires a minimally coherent, more or less consensual framework within which the results of policy issues can be evaluated and the findings of investigation can be interpreted."(8) Second, such controversies threaten to overwhelm societies ability to manage conflict, and so threaten the stability of liberal democracy itself. Controversies are increasingly "settled" (though not resolved) by appeal to the courts, or by policy stalemates which simply maintain the status quo.

Schools of Policy Analysis

Schon and Reid argue that traditional approaches to policy analysis can neither account for intractable controversies, nor aid in their resolution. They review the three main schools of policy analysis, each of which is characterized by its own notion of agent rationality. The authors describe the earliest approach as the policy-analytic approach, and credit Harold Lasswell with its founding.(11-3) This approach focuses on rational choice models of agency, and draws heavily from economic models. Policies are evaluated in terms of their cost to benefit ratio. This approach tends to assume that there is an objectively best solution to any policy problems. This approach to policy analysis rests "on a conception of economic rationality according to which policy problems are seen as instrumental in nature, and policy makers are seen as rational to the extent that they do the best they can do to satisfy the combined welfare functions of those affected by their policies.

This approach suffers from three difficulties. First, it has not been able to establish adequate standards for evaluating policy failure or success. Second, by focusing on evaluating outcomes it neglects processes, and so yields little understanding of why a particular policy succeeds or fails. Third, such analysis tends to be geared toward providing information for legislative oversight, rather than toward providing information for the policy consumers and public.

Shortcomings of the policy-analytic approach sparked the development of the political approach to analyzing policy.(13-6) The political approach emphasized the pluralism, and the presence of conflicting and competing interests. On this view agents are rational insofar as they rationally promote their own interests and values. Interest groups and agents bargain and exchange to achieve the policy or compromise which best satisfies their individual interests.

The authors identify two difficulties with this approach. First, it cannot account for intractable conflicts. Why should some conflicts be immune to bargaining and compromise? Second, it does not address the issue of disproportionate power. Powerful groups may be better able to satisfy their interests at the expense of relatively powerless groups.

The consensual dispute resolution, or mediated negotiation approach draws on its predecessors.(16-9) It seeks to identify policies which best satisfy the welfare of all affected, taking into account the presence of different and possibly competing interests among the parties. Key to achieving such satisfaction among all the parties is to distinguish between the partiesí explicit positions, and their underlying interests. Parties are rational insofar as they attempt to "achieve joint gains for the participants by converting win-lose to win-win situations."(17)

Again, there are two difficulties with this approach. First, some theorists concede that this approach may not be effective in disputes over basic rights or values. That is, it may not be useful in addressing the sorts of policy controversies in which the authors are most interested. Second, it assumes that the parties interests are constant or given. As Schon and Reid ask, "How can one develop reliable approaches to the achievement of joint gains if the participantsí views of gains become unstable?"(19) And yet, some negotiation theorists have claimed one of the greatest benefits of mediated negotiation lies in the partiesí transformed understandings of their own and othersí interests. Indeed the resolution of values conflicts seems to hinge on just such transformations.
 

Chapter Two: Policy Controversies as Frame Conflicts

In this chapter Schon and Reid begin to develop their own approach to policy making, the frame-critical approach. They begin by introducing the concept of frames.

  • Frames: structures of belief, perception, and appreciation which underlie policy positions (23)
  • Policy controversies: disputes in which the contending parties hold conflicting frames (23)

Real situations are complex, vague, ambiguous and indeterminate. In order to make sense of any situation one must select out certain features and relations which are taken to be the most relevant characteristics of that situation. These features allow one to create a story which explains the situation. The authors refer to this selection process as the process of "naming and framing"(26) They explain, "From a problematic situation that is vague, ambiguous and indeterminateÖeach story selects and names different features and relations that become the ëthingsí of the story -- what the story is aboutÖEach story places the features it has selected within the frame of a particular context."(26) Schon and Reid refer to these underlying contexts as generative metaphors. Examples of such generative metaphors include the metaphor of disease and health, of sin and redemption, or of natural processes versus unnatural interference. Framing is necessary to make a problematic situation intelligible. However most situations can be framed in different, and even incompatible, ways.

Frames do more than simply describe a situation. Frames have normative implications, that is, they imply that a certain type of solution is called for. For instance, a problem framed in terms of disease calls for a rather different response than one framed in terms of sin. The authors note that "It is typical of diagnostic-prescriptive stories such as these that they execute the normative leap [from describing a problem to recommending a solution] in such a way as to make it seem graceful, compelling, even obvious."(26) Furthermore, "This sense of obviousness of what is wrong and what needs fixing is the hallmark of policy frames and of the generative metaphors that underlie them, and it is central to our account of the intractability of the frame conflicts implicit in policy controversies."(28)

There is a reciprocal relationship between partiesí interests and the way that they frame a problem. On the one hand their understanding of their interests may motivate them to frame a situation in a particular way. On the other hand their framing of the situation affects their perception of their interests.

Because of this relation between frames and interests, and because of our tendency to interpret facts differently in light of our frames and interests, it is not possible to falsify a frame. It is generally not possible to conclusively disprove or disconfirm someoneís framing of an issue. And so frames cannot be said to be objectively right or wrong, objectively correct or incorrect. In addition, there is no objective stance from which to evaluate frames, if by objective we mean frame-neutral.

Key concepts

The authors offer a number of key concepts and distinctions which they will use to further explore the role of framing in intractable policy controversies.

  • Policy discourse: a verbal exchange or discussion about policy issues (31)
  • Policy forum: the institutional setting or vehicle for policy discourse. Examples include courts, legislatures, political parties, editorial pages, radio and television programs. Different forums have different rules of discourse (32)
  • Rhetorical frame: underlies the persuasive use of story and argument in policy discourse (32)
  • Action frame: underlies policy practice ad implementation. A single policy may have markedly different rhetorical and action frames (32)
  • Policy frame: action frame that an institutional actor uses to construct the problem of a specific policy situation (33)
  • Institutional frame: a more generic action frame from which institutional actors derive their policy frames. The frames that affect how institutional actors frame policy issues. Such frames tend to be complex, and may involve elements typical of the institution, and elements particular to the individual actor. (33)
  • Metacultural frame: broad, culturally shared systems of belief. These frames shape institutional action frames and rhetorical frames. The generative metaphors noted above are examples of metacultural frames. Nature versus nurture debates are another example of contrasting metacultural frames (33-4)

Difficulties in recognizing and reconstructing frames

Framing beliefs are usually not explicitly recognized; they are generally tacit beliefs and assumptions. In attempting to identify and reconstruct (or as the authors say, construct) the various frames that underlie policy issues we face a number of problems.

The first problem the researcher faces in constructing a frame is that it can be difficult to tell which frame is underlying a particular policy position. For example, the rhetorical frame used to "sell" a policy may differ from the action frame that guides its implementation.

A second problem is that the same set of actions may be consistent with very different frames, and so one cannot tell which frame is being relied upon by the actions alone.

Third, different levels of policy administration and application may employ different frames. The meaning of a policy may differ, for example, from the legislature which formulates the policy to the federal agency which enforces it, and differ again at the local level of implementation.

Another difficulty arises in distinguishing between conflict that arise within a frame, and conflicts that cut across frames. This task is made more complicated by the different, nested, levels of frames. A conflict may cut across institutional frames, yet have a shared metacultural frame.

Yet another difficulty arises in trying to distinguish between the potential for shifting a frame, and he actual occurrence of a shift in framing.

A final, more theoretical problem is that the individuals seeking to uncover and construct frames are themselves investigating from within some frame. There seems to be no "unbiased" position from which to reconstruct otherís frames.

Despite these difficulties, Schon and Reid believe that the project of developing a frame-critical approach to policy analysis is feasible. They are hopeful that many of the practical difficulties noted above can be overcome by "carefully nuanced observations and analyses of the processes by which policy utterances and actions evolve over time and at different levels of the policy making process."(36) The validity of frame constructs can be tested against relevant data -- debate transcripts, for instance -- although the authors concede that what the "relevant data" consists of may itself be an object of contention on occasion.

Chapter Three: Rationality, Reframing and Frame Reflection

A central thesis of Schon and Reidís frame-critical approach is that "human being can reflect on and learn about the game of policy making even as they play it, and, more specifically, that they are capable of reflecting in action on the frame conflicts that underlie controversies and account for their intractability."(37) The authors believe that people possess and can develop a frame-critical rationality which will allow them to see how their actions and beliefs can contribute to either the continuation or resolution of policy conflicts.

Schon and Reid begin by considering three conceptual obstacles to developing an account of frame-critical rationality. First is the relation between frame reflection, reframing, and conflict resolution. The second obstacle is the problem of relativism. And the final obstacle is the practical problem of reflecting across frames.

The first difficulty lies in clarifying the relation between reflecting on frames, reframing issues, and resolving controversies. The authors note that frame reflections does not always lead to reframing, and that reframing does not always lead to resolution. Moreover reframing may happen even without any explicit reflection on frames. Frame-critical analysis may not be a panacea for all policy controversies.

The second obstacle that they note is the problem of relativism, which has been alluded to above. The relativist worry is that, in the absence of some frame-neutral standpoint real evaluation of frames is not possible. The validity of any frame must always be relative to some other frame, the validity of which is itself only relative. It would seem then that there can be no final standards or independent criteria from which to judge.

In general, theorists have employed two strategies for dealing with relativism. The first coping strategy appeals "not to a shared perception of fact, but to consensual, logically independent criteria for evaluating frames and choosing among them."(43) The second strategy seeks to translate between different frames. Parties to a conflict would seek to put themselves in the other parties shoes. The authors explain that "in order to do this...each party would have to be able to put in terms of his or her own frame the meaning of the situation as seen by the other in terms of the otherís frame."(45) Reciprocal translation would allow for mutual understanding and could facilitate joint resolutions and the creation of shared frames.

Each strategy raises its own further difficulties. The authors observe that, "the difficulty with any model of frame choice based on superordinate criteria is that the sponsors of conflicting frames are likely to apply the ësameí criteria -- beauty or utility, for example -- in different ways."(44) We will be faced with the further problem of reconciling our understanding of the criteria. And so attempts to employ this strategy lead us to the strategy of reciprocal translation.

The strategy of reciprocal translation presents its own difficulties. Why would conflicting parties attempt the difficult task of discourse across frames? And how could such discourse occur? Schon and Reid argues that the parties BATNAs would be the source of their motivation, if any, to discourse. For models of how such frame crossing discourse might occur they appeal to Thomas Kuhn and Jurgen Habermas.

Kuhn explains how scientists operating from within different paradigms might learn to communicate: "Each may...try to discover what the other would see and say when presented with a stimulus to which his own verbal response would be different...each will have learned to translate the otherís theory and its consequences into his own language and simultaneously to describe in his language the world to which that theory applies."(47) Kuhn suggests that one party may "go native" and adopt the new language as her own. However Kuhn does not specify why or when such conversions would occur.

Habermas theorizes the ideal discourse setting. Discourse should be governed by norms of freedom, openness and justice. Discourse should be free from domination. All parties must have equal rights to raise issues, ask questions, give reasons and present arguments. Parties must be open to accepting or rejecting arguments based solely on the argumentís merits. In practice Seyla Benhabib has argued that these norms demand that we attempt to reverse perspectives with others, that is, that we try to see things from the otherís perspective.(49)

The third obstacle to developing an account of frame-critical rationality lies in implementing reflection across frames. Kuhn and Habermas both give very abstract accounts of cross-frame discourse. In attempting to make these abstract accounts practically applicable, the authors turn to the work of John Forester, Charles Lindblom, James March and Albert Hirschman.

Forester suggests that mediatorsí practice might better approximate ideal discourse if it revised its notions of mediator neutrality. Mediators must acknowledge that when their practice is most effective it does affect the partiesí interests. The goal then would be for mediators to be "íless like experts, judges or implausibly neutral bureaucratsí and more like ënew friends who can create a space for speaking and listening, for difference and respect, for the joint search for new possibilities, and ultimately for newly fashioned agreements about how we shall live together.í"(51)

From Lindblom the authors adopt the idea of disjointed incrementalism, which sees policy analysis as, "incremental, exploratory, serial, marked by adjustments of ends to available means, and socially fragmented,"(52) They also stress the effect that policy inquiry have in unblocking controversies. They authors differ from Lindblom in that they focus their analysis more on specific policies and processes.

March emphasizes the ways in which institutions take on a life of their own. Institutions change in ways that are "complex, uncertain and discontinuous," and that are independent of any individualsí intentions or ability to control.(54) Yet March also points out those distinct and limited spaces in which intentional human action may have an institutional effect.

Finally, The authors adopt Hirschmanís "bias toward hope," namely the hope that "societal development and institutional reform may be deliberately pursued through the exercise of human reason."(55) They also share his goal of producing knowledge which is useful to the practitioner.

Informed by these authorsí views, Schon and Reid develop their own account of frame reflective policy analysis by examining three policy cases. In each case they consider how controversies emerge, how policy positions are reframed, what is the relation of frame reflection to reframing and reframing to resolution, and how policy practitioners reason when they are reasoning well.

Part III: Toward Frame-Reflective Policy Conversation

Chapter Seven: Design Rationality Revisited

Following their examination of the cases, Schon and Reid conclude that "competent practitioners can reflect on the meaning of the policy-making game from a position within it" and argue further that "policy controversies are frame conflicts that may be pragmatically resolved by reframing, and that such frame reflection is central to design rationality--the kind of limited reason that is feasible and appropriate in policy making."(105) In this chapter Schon and Reid present a fuller account of design rationality. They use a number of terms introduced within the case studies.

  • Policy dialectic: a drama in which contention among institutional actors in a policy arena combines with the actorsí adaptations to shifts in a larger policy environment so as to effect, over time, the transformation of a policy object. Policy dialectic is an ongoing process wherein solutions to one problem generate new problems.(81)
  • Design rationality: approaches policy making as a process of design, subject to a design rationality. The policy designer notices how meanings, criteria and constraints change as the policy object takes shape, is attentive to unintended consequences of her policy making, and revises her framing of the policy problem and invents modifications of the policy objects in light of these insights.(86)
  • Policy conversation: a metaphor for policy design seen as communicative interaction between designers and those who use or have a stake in the policy object.(122)
  • Back-talk: in the policy conversation, messages sent back to policy designers that surprise them by violating their taken-for-granted assumptions and tacit "action frames" telling them that the policy object as perceived and used by other actors is different from the one they intended to design and deploy.(123)

Policy design is a complex process. Schon and Reid begin by describing design rationality as it is employed at different levels of complexity. The simplest level involves the designer, the materials, and the intended object. The authors describe the design process at this level:

A designer works with materials to produce an intended object and discovers that the materials resist , more or less, his attempts to impose his intentions upon them. In this process, the designerís intentions evolve. Design moves inevitably produce some unintended effects, which the designer may see as flaws to be corrected, or as happy accidents that suggest new opportunities. In designing, as distinct from instrumental problem solving, something is being made under conditions of uncertainty and complexity, so that it is not initially clear what the problem is or what it would mean to solve it.(166)

Three basic norms of design rationality come into play at this level.

  • The designer must set up the problem in a way that allows for a solution.
  • The formulation of the problem must take into account all of the important constraints, possibilities and uncertainties inherent in the situation.
  • The designer must observe the consequences (intended and unintended) of her designing moves and reformulate her problems and solutions in light of her evolving understanding of the evolving situation.

Another layer of complexity is added when we consider not a lone designer, but multiple actors designing, implementing and consuming policy. Schon and Reid call this level of complexity the communicative-political design drama. Design at this level is a social process in two senses. The lone designer is replaced by the design system with multiple actors. The design object is presented to a broader audience. Design at this level is communicative both because it requires communication among the designers, and because it requires larger policy conversation with policy consumers and other stakeholders. The process is political both because the design system is made up of actors with their own (potentially divergent) interests and power, and because of the interaction with actors in the larger environment.

Rationality in the communicative-political design drama starts with the basic norms. But with the added complexity "the objects of reflection expand to include the designing systemís communication with other actors in the policy drama: the messages sent and received, the interpretations constructed for them, and the tests of such interpretations."(169) The authors suggest additional norms to guide the designers reflection on their communications.

  • Members of the design system should seek agreement on the nature of the problem and the general character and content of a solution. They are attentive to the meanings of the messages they receive, and to the interpretations of the messages they send, and seek to reach a common understanding of these meanings.
  • Designer must gauge their moves to meet both the "substantive requirements of problem-setting and -solving, an the requirements, political and interpersonal, of sustaining the design coalition."(170)

A third layer of complexity is introduced by actorsí conflicting action frames. The design rational response to such frame conflicts is co-design, "in which the contending parties become members of a reformed designing system for the purpose of redesigning the policy objective."(170) Co-designing requires that participants reflect on frames in three ways. They must reflect on and possibly revise their framing of the problem. They must reflect on how the various parties intentions and actions have caused the policy making process to become blocked. They must reflect on the framing of the policy object, and attempt to understand the intentions and meanings of the other parties.

Co-design calls up further norms of design rationality.

  • Designers must reflect on the meanings that underlie back-talk, in order to become aware of any controversies that might necessitate policy redesign or reframing.
  • Designers must also reflect on how their own actions may have contributed to the controversy, or may have precipitated the back-talk.
  • They must reflect on background learning and on facts which have been uncovered over the course of the policy making process.
  • When blocked by controversy, designers must reflect on the structure and politics of the policy process, seeking to unblock the process by marketing and revising the policy object, by negotiation and compromise, or by co-design.
  • Designers may need to invent new elements for the policy object which synthesize aspects of the conflicting action frames.

Schon and Reid distill four basic features which characterize design rationality across levels of complexity.

  • There is a design process which involves making something out of available materials in a complex and uncertain situation.
  • The designerís intentions must always be emerging and evolving in response to the emerging and evolving situation.
  • The designer is engaged in a conversation in which he "seeks to grasp the meanings of his moves, and of otherí responses to his moves, and to embody his interpretations in the invention of further moves."(172)
  • The process of problem-setting and problem-solving can be evaluated by how well it satisfies the partiesí emerging interests, values and intentions, and by what further features of the design situation it discovers.

Contrary to the concerns raised in the Introduction to this book, the authors argue that frame reflection does not require the thinker to distance or disengage herself from the problematic situation. Case studies show that practitioners can engage in situated reflection. They can reflect, for instance, on metacultural frames via the medium of the concrete situation at hand.

In practice, the design rational approach also avoids the relativist problems described in Chapter Three. The cases show that, despite policy participantsí different frames, certain brute facts often emerge for all participants. Differing frames may lead the participants to interpret the causes, implications and importance of such facts in different ways, but the emerging facts are nonetheless the same in some basic sense for all the participants. As the participants each adjust their frames to accommodate the new facts their frames are likely to begin to take on a family resemblance. Shared perceptions and similar experience also aid participants in reflecting on and understanding each otherís frames.

Situated reflection and controversy resolution

One of the authors main claims is the policy controversies which seem intractable in principle may nevertheless be resolved at the practical level by the use of design rationality and situated frame reflection. They offer a number of reasons why progress may be possible at the practical level, when it would seem to be blocked to a more distanced viewer. First, participants tend to be very strongly motivated to make something happen. Second, concrete situations are rich in information, and this informational richness may open up more opportunities for progress. Third, action frames are usually complex and made up of many elements. In practice parties may reach mutually acceptable resolutions by merely shifting the emphasis placed on various elements of their existing action frames, and so without the need for fundamental reframing. Fourth, changes in the situationís larger context may open up new opportunities for resolution. Finally, the presence of others with differing views compels reflection on their perceptions and interpretations. The need for coordination within the design team strongly motivates them to achieve effective communication across their differing frames.

Policy design is a cooperative process, and so its success depends crucially on maintaining mutual trust among the participants. Trusting in such situations means being "prepared to act as though your counterparts will behave cooperatively in spite of the risk that they might not do so and in advance of evidence that reveals how they will behave."(179) Civic spirit, friendship, mutual respect, common values and shared purposes all help to support mutual trust. In order to maintain mutual trust actors must also have the behavioral and communication skills needed to effectively convey their intentions and to express their reliability.

The applicability of a design rational approach

Schon and Reid offer four reasons to believe that design rationality will be generally applicable, that is appropriate and helpful in the great majority of policy situations. First, everyone involved in policy design has a rational interest in maintaining effective communications. Frame conflicts are a source of miscommunication. And so policy designers have a rational interest in frame reflection in order to address frame-based miscommunication.

Second, all policy designers must deal with transactional effects between their actions and their environment. Humans both shape their actions in response to their environment, and shape their environment by their actions. Policy designers "have a rational interest in discovering how their own beliefs may have lead them to undertake actions that helped to create the environmental conditions by which they are constrained." In other words, policy designers have a rational interest in understanding their own action frames.

Third, policy makers have a rational interest in detecting and understanding flaws in their designs. They must try to make sense of peopleís back-talk, protests, unintended uses or disuse of policy objects. In order to make sense of such feedback, policy makers must "be able, again, to put themselves in [other peopleís] shoes, entering into their ways of framing the policy situation and constructing meaning for the policy object."(185) Policy makers have a rational interest in understanding otherís frames.

Finally, hybrid frames created to resolve policy controversies may not be internally consistent. The new frame may contain older elements which, if reactivated or reemphasized, would renew the controversy. Or the reframing may resolve the immediate issue but leave a deeper frame disagreement intact. Policy makers need to understand these framing issues in order to seek more lasting and stable resolutions.

Chapter Eight: Conclusion, Implications for Research and Education

The split between reason and practice noted in the Introduction also shows up in policy research and education. The rigorously tested truths of the social sciences tend to be trivial and of little use to the practitioner. On the other hand the heuristic principles of craft distilled from examination of successful practice tend to be "mutually contradictory, noncumulative, not generally applicable, and nontestable."(191) Schon and Reid hope that this split can be healed by emphasizing design rationality in research and education.

Policy research must address policy practice. "Policy researchers should focus on the substantive issues with which policy makers deal, the situations within which controversies about such issues arise, the kinds of inquiry carried out by those practitioners who participate in a controversy or try to help resolve it, and the evolution of the policy dialectic within which practitioners play their roles as policy inquirers."(193) Research by policy academics should be carried out in collaboration with policy practitioners.

The authors suggest a number of ways in which such collaboration could occur. They might analyze success stories, study current policy issues, study past policy controversies, or engage in frame reflective studies of ongoing policy design processes. Collaboration between academics and practitioners would have two main benefits. It would help in the process of reciprocal frame reflection. And it could help practitioners to maintain the conditions needed for mutual trust. Research of this type could be thought of as a type of frame-reflective consultation. The test of such research would be how useful it was to policy practitioners.

When policy research does not focus on practice it can exacerbate tensions within the field. The authors note that there are distinctions among policy practitioners. There are the practitioners who design policy, and practitioners who are more involved in implementing established policy. There is often some tension between these two branches of practitioners. Designerís interests in accountability in implementation are often in tension with field workers desires for autonomy. Policy researchers may exacerbate this tension. Policy researchers are often hired by oversight agencies to evaluate policy programs. "Such evaluative research is not designed to help practitioners perform better but to evaluate whether their activity is worth doing in the first place." Since such evaluations are very often program killers, practitioners often find policy research threatening.

For this reason the authors recommend a triadic approach to research collaboration. Research should involve collaboration between academics, designers, and every-day practitioners.

The research setting should be neither so close to the policy situation as to completely embroil the researcher, nor so distant as to leave the researcher disconnected from the situation. Researchers should seek an optimally distant setting, "close enough to the actual policy situation so that a practitioner move in and out of it with ease, yet distant enough from it so that she is protected from its pressures, threats, and distractions."(199)

Practice oriented policy research must have standards of generality, rigor and validity which are appropriate to it. The causal model used in much of the social sciences (which seeks the unique determinants of an event) in not appropriate to policy research. Policy research should instead emphasize the causal tracing of events, generating causal stories. Such causal stories are validated to the extent that they solve the original problem, just as the success of a medical treatment may serve to confirm the original diagnosis. While normal social science attempts to formulate covering laws, policy research should instead aim to describe generalizable patterns. Such patterns are employed by practitioners via the process of reflective transfer, which is "the process by which patterns detected in one situation are carried over as projective models to other situations where they are used to generate new causal inferences and are subjected to new, situation-specific tests of internal validity."(204)

Policy education

Schon and Reid argue that current policy education focuses on choice and decision-making, and neglects issues of problem-setting and formulation, and of implementation. They suggest that policy education should aim at producing "policy-specific versions of a generic design capacity."(207) Policy schools might model themselves after design schools. In particular they should adopt the idea of the design studio to create a reflective policy practicum as a key part of policy education. In such a practicum students, under the guidance of an instructor, would respond to a problematic policy situation. The practicum should be a virtual policy arena which represents much of the complexity, ambiguity, and evolving meanings of actual policy conversations. Within such a practicum students would have to be learn the task of frame reflection, and to learn how to build and sustain mutual trust.