Amid the diversity of educational standards definitions offered in the recent literature, it is often difficult to discern exactly what standards are or what they are supposed to do.¹ In fact, the rhetoric of standards has often operated as what Michael Apple has called a "slogan system," encompassing "a penumbra of vagueness so that powerful groups or individuals who would otherwise disagree can fit under the umbrella."² This paper aims to promote a more nuanced dialogue in education about standards and how they might be used to orient reform. Drawing on the work of Gary Sykes and Peter Plastrik, we argue that it is useful to see many definitions as lying on a continuum running from a vision of standards as promoting dialogue and democracy, to a vision of standards as transmitting authority and control.³

On the more democratic end of this continuum, standards tend most frequently to be represented by metaphors like "emblems" or "rallying flags."⁴ For example, Deborah Ball, a prominent member of the group that created the NCTM Professional Teaching Standards, notes that "these standards do not comprise proved statements about how to teach mathematics but a set of commitments about good teaching. They were designed to provide a medium for focused and grounded conversation about such teaching, to create a context of direction for the development of practice."⁵ Standards at this end of the continuum appear to be largely intended to reflect the common projects of a community, defining who "we" are and indicating a direction towards which "we" are moving, embodying shared goals and ideals that can be interpreted in multiple ways and participated in by a diverse group.

The "authority" pole of the continuum is also prominent in the literature on educational standards. Those who promote this approach to standards tend to argue, as Diane Ravitch does, that "every meaningful standard offers a realistic prospect of evaluation; if there were no way to know whether anyone was actually meeting the standard, it would have no value or meaning."⁶ Thus, the "authoritative" vision tends to stress the need for standards to provide relatively fixed guidance for reliable judgment, bringing with them a sense of authority and duty, and a power to guide the action of those who are subject to them.

This paper explores two prominent strategies for using standards to reform teaching, teacher professionalization and teacher networks, that lie on contrasting ends of this authority/democracy continuum. Each balances out the competing demands of authority and democracy differently. We suggest that the assumptions about standards that lie behind these projects are usefully reflected in the work of Jürgen Habermas, for professionalization, and Hannah Arendt, for teacher networks. While no theory can completely represent the complexity encompassed by these different approaches, we argue that these two theoretical models give insight into the possibilities and limitations entailed by each. We conclude by speculating on ways the two approaches might inform each other.

Among other goals, teacher professionalization aims to empower veteran teachers, both individually and collectively, while providing clear direction to provisional and relatively novice members of the teaching community. Linda Darling-Hammond, Arthur Wise, and Stephen Klein's book, A License to Teach: Building a Profession for 21st-Century Schools, provides an excellent example. Darling-Hammond, Wise, and Klein agree with many critics of schools that the knowledge and skills of the current teaching force are not high enough. They disagree, however, with the "micro-management" solution generally adopted by governmental agencies in response. While the micro-management response "is understandable given the laissez-faire approach to admitting individuals to teaching," they argue that it is ultimately ineffectual because "the kind of teaching required to meet...demands for more thoughtful learning cannot be produced through teacher-proof materials or regulated curricula" (LT, 9, 2). Instead, teaching resembles the kind of activity engaged in by professions like medicine and law because its "complexity requires non-routine judgments based on an understanding and analysis of many, often competing, considerations" (LT, 13). Thus they argue that effective teaching reform requires not micro-management, but instead a significant improvement in teacher's skills, accompanied by enough autonomy to allow teachers to make non-routine decisions in unpredictable, contingent contexts.

Darling-Hammond, Wise, and Klein's plan for professionalization, then, is modeled on professions like medicine and law in which carefully designed requirements and assessments ensure that all licensed professionals are truly qualified to operate without direct supervision. Importantly, the kinds of assessments they recommend are extremely complex, requiring evaluators to make fine-grained judgments about myriad interconnected pieces of evidence in order to assess the multifaceted, creative skills of professionals (see LT, chap. 4). As in these other professions, Darling-Hammond, Wise, and Klein believe strongly that the responsibility for creating qualifying assessments and for developing the shared standards that orient them should be concentrated in the hands of teachers themselves (see LT, chap. 7, 9). They envision a profession coming together to collaboratively define the characteristics of good teaching - the standards - thereby establishing the characteristics of their collective professional identity. By promoting both spheres of dialogue in which standards and assessments are created, and arenas of evaluation in which teachers are assessed under these standards, Darling-Hammond, Wise, and Klein's model of professionalization balances the competing demands of democracy and authority by splitting the democratic process of standards creation off from the hierarchical and authoritative process of standards application through assessments.

Habermas's recent writings on the creation and application of laws in democratic societies seem to reflect important aspects of Darling-Hammond, Wise, and Klein's approach, providing a theoretical framework for understanding many of its assumptions, possibilities, and limitations.⁷ Habermas seeks a procedure for grounding knowledge in the process of rational argumentation that he argues is inherent in all uses of language. For Habermas, in this type of reflective "communicative action," participants attempt to reach consensus with others by giving reasons and grappling with the arguments presented by their dialogic partners.⁸ Agreements are only valid if they are agreed on because of the power of good reasons, if they are unaffected by coercion or the effects of unequal power, and if everyone has an equal chance to contribute to the dialogue.⁹ Expert, research-based knowledge, he argues, is only qualitatively different from this everyday communicative action because it is "naturally fallible and rarely value-neutral" serving its purpose best when it informs but does not direct or dominate more broad-based everyday discourse about collective social issues (BFN, 164).

In our daily lives, Habermas argues, we operate within "the massive background consensus" of a shared societal "lifeworld," allowing us to live together without constantly engaging in communicative action (BFN, 22). It is only when we have a problem that such reflective communicative action becomes necessary. Any agreements that are reached in dialogue, then, are aimed ultimately at reconstructing "the intersubjectivity of our life-world," at overcoming these disagreements so they cease to be barriers to group activity; thus dialogue always aims at consensus.¹⁰ Importantly, Habermas argues that for a consensus of this kind to be achieved, participants must agree on the same conclusion for the same reasons, or they will each understand the reasons that are given in different ways, and their final agreement will look different to each of the different participants (BFN, 11). Thus, what he calls "aesthetic" approaches to dialogue, in which each person interprets what is being discussed from her own perspective, must be avoided as much as possible.¹¹

Habermas is clear, however, that the requirements of communicative action are never entirely achieved in the real world, and he presents communicative action only as a useful guiding ideal. Because the ideal is never reached, and since the world is constantly changing, agreements reached in contexts that approximate communicative action are always open to reassessment and revision in response to new dialogic contributions, creating a condition of "permanently endangered counterfactual assumptions" (BFN, 21).

In an ideal society, Habermas imagines, every time we made a decision we would speak with those who would be affected by it and come to consensus before any action was taken (BFN, 323-27). He realizes that this approach would be impossibly burdensome even in the smallest community, however. In part, he solves this problem by arguing that every concrete instantiation of communicative action requires local participants to attempt to at least imagine the kinds of reasons those not present might raise (BFN, 14). Habermas acknowledges, however, that truly legitimate decisions would require that the opinions of those not present in a particular context be actually and not simply imaginatively taken into account. Habermas thus posits what he calls the "public" in which the distributed, overlapping dialogues of particular groups of individuals slowly aggregate, over time, into collective "opinions" that are detached from the particulars of local contexts and the individuals who contributed to them. The existence of these "public opinions" makes it possible for committees made up of "the broadest possible spectrum of interpretive perspectives, including the views and voices of marginal groups" to arrive at legitimate decisions (BFN, 183). For this to work, these committees must be informed through many different channels of communication about the range of public opinions, approximating the requirements of an ideal speech situation that includes the relevant contributions of all (see BFN, 170-71).

While this theoretical discussion may seem initially distant from the practical concerns of policy-makers, it actually maps quite well onto the assumptions implicit in many standards documents.¹² These generally report that a range of different streams of dialogic data informed the actions of a broadly representative committee that drew together extensive research data while widely distributing initial drafts of standards documents for public comment.

It is important to note, given this model, that in the context of teaching reform different kinds of standards often seem to be created in different communal contexts. For example, while "content" standards (what teachers are supposed to be able to "know and do") often draw on the widest commentary and dialogue, "performance" standards (the level of achievement qualified teachers are supposed to achieve), generally involve a much more limited community, consisting, in the case of teaching standards, for example, of those most directly involved with the activity of teaching. While informed by the content standards, efforts to create performance standards appear often to represent a shift into a more refined and "expert" dialogue among teachers, something that fits well with at least Darling-Hammond's vision.¹³

Even with the representative committee approach explored above, however, Habermas argues that in complex modern societies it is still impossible for every decision to issue from the time-consuming process of democratic dialogue. Modern societies cannot avoid giving crucial "steering" functions over to systems of power, like the market or bureaucracy, that operate not through fair dialogue but according to their own internal logics.¹⁴ The only way societies can subject these systems to democratic control, Habermas argues in his most recent writings, is through something like a legal system. This is because unlike other steering systems, the legal system remains ultimately subject to the rationally achieved consensus represented by dialogically created laws (or standards in our case). Maintaining the democratic character of the legal system, however, requires that the democratic processes of law creation be largely separated from systems of application (see BFN, 188). In the special case of educational assessment, this indicates that while the standards remain open to critique in general, in the assessment context once they are established they are supposed to serve individual judges as relatively fixed guidelines. Although standards must evolve to some extent in the process of judicial application to particular cases, Habermas does not want assessment practices to "self-program" themselves in ways that exceed the mandate provided by the community consensus that initially created them (BFN, 220-34,188).

This requirement creates problems for assessment systems, however. Habermas is not under the illusion that people can always reach consensus. He is aiming in his work not for a society where we are all the same, but instead for a pluralistic society in which life-plans of great diversity can live together. As the diversity in a society grows, however, standards that intend to represent a variety of groups equally must become increasingly abstract because the more diverse we are as a culture, the less specific will be the issues we can agree on without forcing some to give up something crucial about who they are (BFN, 140). Yet, at the same time, the more finely we intend to discriminate between teaching performances, the more concrete our standards would apparently need to be, and the more our standards would tend to represent one "ethical" "form of life" over another.¹⁵ And the complex nature of the performance assessments recommended by Darling-Hammond, Wise, and Klein would seem to require this kind of fine discrimination.¹⁶ Thus, the consensual standards we can hope to achieve noncoercively in many teaching fields may be fairly abstract, and these abstract standards may have difficulty serving the kinds of defined sorting functions that Darling-Hammond, Wise, and Klein recommend. In fact, Habermas does grapple with this problem of abstraction, presenting some tentative solutions. He argues that applying relatively abstract standards to concrete problems on a local level requires that more dialogic contexts be created on these local levels to allow "further specification" of the standards (BFN, 439). "This implies," he says, "a 'democratization' of the administration that...would supplement parliamentary and judicial controls from within" (BFN, 440).

The problem with the idea of standards-oriented performance assessment, however, goes beyond that of abstraction. As we have noted, fundamental to Habermas's model is the assumption that the standards we agree on in approximations of communicative action have identical meanings for all participants. The more each participant in the process of standards creation interprets the standards they have agreed upon in different ways, the more difficult it becomes to understand how these standards could be used to initiate judges into single perspectives on what counts as "good" teaching. Even if a judge is supplemented by discursive contributions from the local level when facing a difficult decision, in some, perhaps many, cases the multiple perspectives these contributions represent, layered over the multiple perspectives contained in the standards already, cannot be integrated into a clear consensus because the problem a group is discussing is not solved in exactly the same way for each person. The standards they agree on have different meanings for each. As Apple notes, a "penumbra of vagueness," allowing multiple interpretations to exist within commonly "agreed upon" standards is common.

As diversity in a community increases, then, it becomes increasingly important to question how consensually achieved standards might serve the function Darling-Hammond, Wise, and Klein and others desire from them. As Sally Lubeck points out, assessments that treat standards as if they represent an achieved consensus of the community can end up suppressing the diversity of interpretations resident in the field.¹⁷

Yet, despite these problems, Darling-Hammond, Wise, and Klein's vision of a more empowered profession of teachers remains vitally important. The state clearly has an obligation to protect children from "incompetent" teachers in ways that do not reduce the educational efficacy of the entire educational system as micro-management appears to. The question, then, is how we might achieve important aspects of Darling-Hammond, Wise, and Klein's vision while living within the limitations of the kinds of standards we can realistically achieve.

Teacher networks appear to represent an approach to school reform fundamentally different from assessment-based professionalization, drawing from a fundamentally different understanding of "standards." We focus here on the recent work of Ann Lieberman and her colleagues. Lieberman and Maureen Grolnick note that many teacher networks seek to overcome the more hierarchical and bureaucratic structure of schools by "creating 'public spaces' in which educators can work together in ways that are different in quality and kind from those typical of their institutions."¹⁸ To succeed in this task, Lieberman and Grolnick argue, networks must "rally prospective participants to a particular cause, idea, or set of connections."¹⁹ Thus, we will suggest that in network approaches to reform, discursive objects like standards provide groups with common focal points that shift in response to the contributions made by different group members. Whereas standards used in centralized assessment of teachers appear to represent solutions to the problem of what constitutes good teaching, standards used in teacher networks appear more likely to be treated as common issues to be struggled with collectively, creating a shared identity through the interpretive participation of network members.

The democratic aspect of networks should not be overemphasized, however. Like Darling-Hammond, Wise, and Klein's plan for professionalization, networks struggle with the tension between democracy and authority. As Lieberman and Milbrey McLaughlin note, networks cannot survive without some measure of centralization and hierarchy. For example, networks struggle to maintain the "quality" of their pedagogy, leading many networks to depend on complex relationships with sources of "authoritative" expert knowledge, like universities, founding members, or central offices.²⁰ Networks also struggle with issues of leadership and organization. Without leaders who are "visionary, multicultural,...at ease with ambiguity and flexibility,...and able to nurture emergent talent," networks threaten to "become very much like the bureaucracies they are trying to change."²¹ And, with respect to the shared goals of a network, "a consistent problem surfaces: Who controls the agenda?"²² Without some common identity, some limits, however tenuous, on what counts as a "proper" interpretation of their shared goals and ideals, it seems there is no network.

Arendt's theory of "public space" appears to provide a useful theoretical framework for understanding the relationship between "standards" and reform networks as represented by Lieberman and her colleagues. Like Habermas, Arendt imagines dialogic spaces in which individuals might engage in noncoercive dialogue together. Unlike Habermas, however, participants in an Arendtian public space do not aim at the achievement of consensus.

According to Arendt, public spaces come into being when there is what she calls an "in-between" made up of a set of common issues or concepts which "must be seen in a variety of aspects without changing their identity, so that those who are gathered around them know they see sameness in utter diversity."²³ Each individual must be able to arrive at her own interpretation of their common project, yet not interpret it so uniquely that the "common" nature of the object is lost. In networks, we would argue, this "in-between" is represented by the common goals and ideals of the organization, the "standards" that act as what Ball calls a "medium for grounded and focused conversation." Standards in contexts like these appear to take on meaning at least partly through the interpretations given to them by each participant (for example, HC, 183).

In an Arendtian vision of "standards" then, a clear distinction cannot be made between standards creation and application. Each act of "application" is also, at the same time, a recreation of the standard that is being engaged with. Public spaces create contexts in which individuals can continually explore and develop what are to some extent their own particular visions of teaching as a part of a common effort with others. In contrast with Habermas, then, Arendt's vision of standards draws fundamentally from the "aesthetic" interpretive activity of individuals engaging in dialogue. In an Arendtian public space, people take unique yet communicable stances on common issues; the space, therefore, is constituted by the multiple perspectives that constantly contribute to it (HC, 201).

Arendt is quite clear, however, that she conceptualizes the public (unlike Habermas) as a largely local phenomenon. She argues that, as a public space increases in size beyond the point that it can be organized through face-to-face interactions, more bureaucratic forms of organization tend to emerge (HC, 43). Not surprisingly, then, one of the fundamental challenges facing teacher networks is dealing with expansion. As Lieberman and Grolnick discovered, "many [network leaders] spoke simply about being aware of the problem of becoming too big, or too bureaucratized."²⁴ In fact, Arendt continually emphasizes how precarious the achievement of a public space is. Its continued existence depends upon the committed participation of new and old members that will carry forward and build upon the vision embodied in the group's "standards."

From the discussion above, it should be clear that an approach to teaching reform based in the creation of public spaces among teachers has some fundamental drawbacks. For example, limitations arise from the fact that networks are dependent upon voluntary engagement. Networks may often lack the leverage necessary to fundamentally change the hierarchical and bureaucratic practices that currently characterize schools today, and the focused strength that can arise from the more hierarchical bureaucratic organization reflected in the Darling-Hammond, Wise, and Klein model.²⁵ In addition, Darling-Hammond, Wise, and Klein's work suggests that giving up on authoritative uses of the standards, may paradoxically promote the creation of systems of bureaucratic "micro-management" by frustrated governmental bodies. Finally, there is a tension between a desire to include diversity in networks and their need to maintain a sense of coherence and identity. The need for members to recognize the interpretations of other participants as "relevant" contributions to their common project may require groups to draw boundaries, however blurred, between acceptable and unacceptable interpretations, boundaries that appear to often systematically exclude some groups.²⁶

Yet despite their fragility and problems with diversity, public spaces organized around shared standards appear to provide examples of contexts in which democracy and the unique potential of many individual teachers might be nurtured in ways that a more Habermasian vision of reform may have difficulty embracing.

While these two approaches to standards may seem initially incompatible, we will suggest that there are ways that they might work together, although the brief discussion that follows cannot fully capture the complexity of such an attempt. In the context of a large-scale effort to assess beginning-teacher portfolios, for example, we have begun to explore whether evaluators might be trained to recognize portfolios that contain issues on which the larger community is not in agreement. Portfolios identified as "problematic," we argue, might be shifted from a more "authoritative" Habermasian assessment structure to a more Arendtian deliberative context in which a diversity of perspectives is acknowledged. A diverse committee of evaluators might then engage in dialogue about the portfolio and the relevant standards, each evaluator attempting to understand the perspectives of the others. While consensus in this more Arendtian context would not be resisted, difference would be valued as well (and the participants might come to consensus on a "score" for a particular candidate for different individual sets of reasons).²⁷ In fact, drawing from Habermas while rejecting his valorization of consensus, Georgia Warnke argues, that through a process similar to this "in taking seriously the interpretive insights of the other...we can begin to improve our own.... In a conversation of this kind each interpretive stance may retain its distinctiveness; it can also help to develop and enrich the others and, in turn, be developed and transformed by them."²⁸ We have only begun to grapple with how such a committee might reach a decision when they cannot agree, although we would argue there are a range of possible options that might be pursued.

It is important to note that regardless of the system we might develop to deal with any dissensus resident in a particular set of orienting standards, to the extent that the decisions we make eliminate controversial teachers from the pool of "professionals," we invariably affect the makeup of the teacher community that will have primary responsibility for recreating the profession's standards and assessments in the future. If standards are at least partly designed, as Ball and others have argued, to point to an unknown future, we must acknowledge that controversial teachers may just as well represent the not yet understood innovations of tomorrow as the problems of today. Decisions about who will join the profession will have as much of an impact on the nature of our dialogic community as it will on the futures of particular children in classrooms. It is possible that an assessment system that drew on a more Arendtian model of judgement might mitigate this. At the same time, we imagine that a more dialogic approach to assessment might also contribute, along with teacher networks and other forces, to the slow development of a more vibrant democratic "civic" culture among veteran teachers outside schools, something Habermas acknowledges is a vital precursor to the legitimate generation of the standards on which his entire model depends, but which his model cannot generate within itself (BFN, 131).

For a response to this essay, see Wagner.

28. Georgia Warnke, Justice and Interpretation (Cambridge: MIT Press, 1993), 131-32.