The Limits of Dialogue as a Critical Pedagogy

Nicholas C. Burbules

Department of Educational Policy Studies

University of Illinois, Urbana/Champaign


Published in Revolutionary Pedagogies, Peter Trifonas, ed. (Routledge 2000)



It seems that hardly anyone has a bad word to say against dialogue. A broad range of political orientations hold out the aim of "fostering dialogue" as a potential resolution to social conflict and as a basis for rational public deliberation. A range of pedagogical approaches, from constructivist scaffolding to Socratic instruction to Freirean liberatory pedagogy, all proclaim the virtues of an interactive engagement of questions and answers in the shared pursuit of knowledge and understanding. Philosophical accounts of dialogue from Plato to the present employ the dialogical form as a literary genre that represents the external expression of an internal, dialectical thought process of back-and-forth ratiocination. Dialogue constitutes a point of opportunity at which these three interests — political, pedagogical, and philosophical — come together. It is widely assumed that the aim of teaching with and through dialogue serves democracy, promotes communication across difference, and enables the active co-construction of new knowledge and understandings.

Nevertheless, the ideal of dialogue has received withering criticism, particularly from poststructural feminist theorists in education and from those for whom "difference" is a lived experience of marginalization, and not just a demographic category of identification. For these critics, "dialogue" has exerted a kind of hegemonic dominance that belies its emancipatory rhetoric, its apparent openness to difference, and its stress on equality and reciprocity within the dialogical relation. The way in which dialogue has become almost synonymous with critical pedagogy has tended to submerge the voices and concerns of groups who feel themselves closed out of dialogue, or compelled to join it only at the cost of restricting their self-expression into acceptable channels of communication. Finally, an idealized, prescriptive conception of dialogue has abstracted the situated historicity of specific practices of communicative engagement from their consequences for people and groups who encounter the invitation to dialogue in difficult circumstances of conflict.

In light of such reactions, the claims made on behalf of dialogue as an inherently liberatory pedagogy need to be reassessed. The insistence that dialogue is somehow self-corrective, that if there are unresolved power differentials or unexamined silences and omissions within a dialogue, simply persisting with the same forms of dialogical exchange can bring them to light, seems not only counterproductive but itself a form of hegemony: If dialogue fails, the solution to the problem is more of the same.

Yet it also remains true that the ideal of "dialogue" expresses a hope in the possibility of open, respectful, critical engagements from which we can learn about others, about the world, and about ourselves. Is there a space between the exaggerated claims made on behalf of dialogue as an inherently liberatory pedagogy and the rejection of dialogue as an ideal entirely? Can dialogue continue in good faith while acknowledging the inherent limits to (and dangers arising from) its aspirations toward understanding across differences? Or must such aspirations toward understanding and communication be abandoned entirely? These are the questions animating this essay.

The Fetishization of Dialogue

We seem to be living in an era in which for many "dialogue" has become the foundation of last resort in an antifoundational world. The thoroughgoing proceduralism of placing trust in processes of interpersonal communication has proven to be compatible with a wide range of otherwise quite different social and political stances. Dialogue represents, to one view or another, a way of reconciling differences; a means of promoting empathy and understanding for others; a mode of collaborative inquiry; a method of critically comparing and testing alternative hypotheses; a form of constructivist teaching and learning; a forum for deliberation and negotiation about public policy differences; a therapeutic engagement of self- and other-exploration; and a basis for shaping uncoerced social and political consensus. I will briefly review six dominant traditions that have centrally invoked the concept of dialogue, particularly in relation to the aims and methods of education.

(1) For liberal views of dialogue, such as those of John Dewey or Benjamin Barber, dialogue is the fulcrum around which the imperatives of democracy can be reconciled with the facts of diversity and conflict. For exponents of "deliberative democracy" it is in public, communicative engagements that democracy works its will, and a chief aim of democratic education must be to foster in learners the capabilities and dispositions to participate in such deliberations. An implication of this stance, however, is that those who do not, who cannot, or who choose not to develop or exercise these capabilities suffer an attenuated relation, at best, to the democratic public sphere, if not an actual exclusion from it: "Proponents of liberal dialogue are not sensitive enough to the fact that a theory of conversational restraint may be damaging precisely to the interests of those groups that have not been traditional actors in the public space of liberalism — like women, nonwhite peoples, and sometimes nonpropertied males." Public education is supposed to be an arena of training for engagement in the rough and tumble of public deliberation; but the very avenues of opportunity for access to deliberation on these terms can be seen from a different vantage point as barriers of exclusion.

(2) Some versions of feminism, by contrast, tend to reject the agonistic features of dialogue in this sense and to promote a more receptive, caring stance in the dialogical relation. Deborah Tannen’s recent popular book, The Argument Culture: Moving from Debate to Dialogue, is an extended paean to this nonconfrontational view. More detailed and modulated treatments of this theme can be found in Mary Belenky et al., Carol Gilligan, and Nel Noddings. What relates all of these accounts is a linkage between a competitive, adversarial approach to public or private disagreements and the stereotypical norms of masculine behavior, and the association of "dialogue" with the more open, receptive, inclusive spirit of women’s values. Educational and social deliberation that privileges the more adversarial mode of interaction, and discourages or dismisses the more tentative and cooperative spirit of dialogue, on this view, discriminates against females in schools and in the public sphere generally. These authors are always careful to insist that this more receptive stance does not preclude vigorous disagreement and self-assertion, but it is not difficult to see why these views have come to be labeled by other feminists as "good girl" feminism. Without intruding myself into this particular disagreement, I think it is clear why the mode of dialogue proposed under this view of feminism has not been seen as adequate for the more confrontational politics favored by certain other feminists and by the aggrieved members of other groups.

(3) Platonic views of dialogue stress the role of communicative interchange as a proving ground for inquiries into truth: While in his dialogues the protagonist Socrates distinguishes "disputatious" and "friendly" forms of dialogue (paralleling in some ways the distinction just explored under (2)), in both forms the joint endeavor is to propose and oppose, to formulate arguments and to put forth counterexamples and counterarguments, as the mechanism by which truth is ascertained. It is an intriguing feature of this view, reflected later in a different context in the work of Freire and others, that this philosophical conception of dialogue coincides with a preferred pedagogy: for in Plato’s view the way in which knowledge claims are adjudicated and tested is also the way to teach. Dialogue is a way of drawing forth latent, unformed understandings and facilitating the discovery of truths by the learner for himself or herself — hence the ubiquity of teachers from law schools to kindergartens to adult literacy programs ascribing their teaching to "the Socratic method" (though this method never comprised only one style of teaching). But the Platonic view of dialogue rested upon a view of knowledge as absolute, unchanging, and humanly attainable through recollection — an epistemological stance that almost no one would feel comfortable with today. I suspect that few contemporary advocates of the Socratic method as a pedagogy would want to be held to the underpinnings by which Plato advocated and justified it.

(4) Hermeneutic views of dialogue tend to emphasize dialogue as a condition of intersubjective understanding: what Hans-Georg Gadamer calls "the fusing of horizons." A precursor of this view can be found in the existential theology of Martin Buber’s I-Thou relation. Hermeneutic dialogue emphasizes the relational, to-and-fro movement of question and answer as an avenue toward understanding and agreement. This intersubjective confirmation stands in direct contrast to the objectivist view of convergence around the truth that we find in Platonic views of dialogue. But critics of this hermeneutic view of dialogue have tended to question its limited capacity for critique and for engaging issues of power and inequality that stand outside the dialogical relation; these contexts need to be problematized in terms that go beyond their impact on interpersonal understanding. Moreover, some have questioned the aim of understanding itself as insufficiently attuned to cultural differences and as dangerously naive in supposing that when "fusing" occurs it occurs on neutral ground:

By communicative dialogue, I mean a controlled process of interaction that seeks successful communication, defined as the moment of full understanding. For those who advocate it in education, communicative dialogue drives toward mutual understanding as a pedagogical ideal….What kind of knowledge does dialogue proffer? What techniques does it use to regulate knowledge and the relationship of the teacher and student within the dialogue to knowledge and truth? I’m persuaded that dialogue…is not just a neutral conduit of insights, discoveries, understandings, agreements, or disagreements. It has a constitutive force. It is a tool, it is for something….[It] tries to accurately represent the world through the conventions and politics of realism.

(5) Most contemporary critical views of dialogue, especially those in education, invariably refer to the important work of Paulo Freire. Indeed, for an entire generation of critical educators his writings and life’s work stand as an inspiring model of committed pedagogy, and he has had a primary impact on the work of widely read North American authors including Henry Giroux, Peter McLaren, and Ira Shor. Yet it must be said that this very popularity and loyalty have interfered at times with the selective, critical appropriation and reinterpretation of his ideas. Freire’s distinctions of monological vs. dialogical pedagogies, his critique of "banking" forms of education as the mere "depositing" of information in the minds of students, his conception of conscientization as the overcoming of what he calls "intransitive consciousness," are all virtually canonical. Freirean pedagogy is sometimes taken as simply synonymous with critical pedagogy or radical pedagogy, forcing feminists and others to find different ways of describing alternative critical educational theories and practices. Yet, as the roots of Freire’s pedagogy have come to be more clearly identified in specifically Hegelian, Marxist, and Catholic assumptions, it has become necessary to ask whether this particular constellation of theories is the best or only basis for a radical theory and practice of pedagogy. In some accounts, Freirean dialogue is regarded as a practice with intrinsic critical and emancipatory potential; but many authors, notably some feminists, do not find space within it for critique and emancipation on their terms.

(6) Finally, there are what might be termed post-liberal views of dialogue, especially the work of Jurgen Habermas. Perhaps no contemporary theorist has gone further in proposing a model of communicative dialogue as the nonfoundational foundation for epistemological, political, and moral adjudication. For Habermas, all claims are filtered through the medium of discourse, but it is a medium with evaluative standards built-in: Communicative claims rest upon implicit norms that can be, and should be, critically questioned and redeemed. The grounding of truth and value claims lies in the uncoerced consensus that such deliberations can achieve — including, significantly, critical reflection on the conditions under which that agreement is obtained. These conditions — uncoerced consensus and the implicit norms (discursively redeemed) that regulate communicative interactions — give the outcomes of such deliberation a generalizability not based upon absolute claims of truth or rightness, but secured on the nonrelative criterion of valid agreement among those parties concerned.

Critics of Habermas, including Seyla Benhabib, have complained that this account of communication assumes a commonality in modes of communication and a kind of impersonality toward the way in which participants in deliberation are identified: the emphasis is on the conditions under which consensus is obtained, not the specific choices and identities of those party to it. While sharing the basic idea of discursive justification, Benhabib wants to situate this process in the actual identities, positions, and differences among participants. She calls this "interactive universalism":

Interactive universalism acknowledges the plurality of modes of being human, and differences among humans, without endorsing all these pluralities and differences as morally and politically valid. While agreeing that normative disputes can be settled rationally, and that fairness, reciprocity, and some procedure of universalizability are constituents, that is necessary conditions of the moral standpoint, interactive universalism regards difference as a starting point for reflection and action. In this sense, "universality" is a regulative ideal that does not deny our embodied and embedded identity, but aims at developing moral attitudes and encouraging political transformations that can yield a point of view acceptable to all. Universality is not the ideal consensus of fictitiously defined selves, but the concrete process in politics and morals of the struggle of concrete, embodied selves, striving for autonomy.

Benhabib’s move, emphasizing the actual difference, embodiment, and situatedness of communicative participants, and her shift from rational agreement per se to the ongoing conditions (social and interpersonal) that can support sustained deliberation among contesting points of view, make the Habermasian model both more concrete and more responsive to the fact of cultural diversity.

Nevertheless, even this account has been challenged, for example by Judith Butler, as insufficiently sensitive to difference and as essentially normalizing, that is, tending to discipline the acceptable forms of communication in terms of dominant norms. For Butler and other poststructural critics, the process of relentlessly problematizing conventional norms and categories proceeds through the interrogation of the silences, gaps, and paradoxes of inclusion/exclusion that bedevil even the most "participatory" models of public deliberation — including the disciplinary regimes that suggest (however invitingly) "we fully welcome your participation on these terms." The subtle workings-out of asymmetries of power and access often belie the open and reasonable self-conception of the Habermasian (or even the Benhabibian) models of communication, making even their sincerely invitational gestures a kind of false seduction into conformity. For Butler, Ellsworth, Lather, and other critics, the response is to resist the "good behavior" that is made a condition of participation in favor of what Anderson calls "performative subversion," pointedly refusing to valorize such conditions.

These six conceptions of dialogue comprise almost the entire range of discussion about the topic within the field of education. I have briefly reviewed them, and some of the prominent criticisms against them, not to engage each of these debates in detail, but to draw the background against which current disputes over dialogue are situated. While these six views are quite different from one another, and indeed disagree among themselves over many issues, they have certain crucial features in common. They all place primary emphasis on dialogue as the adjudicative basis for social and political discussion and disagreement. They all privilege dialogue as the basis for arriving at valid intersubjective understanding or knowledge. And they all, in the educational domain, recommend dialogue as the mode of pedagogical engagement best able to promote learning, autonomy, and an understanding of one’s self in relation to others. The prominence of these six views, particularly among educational theorists and practitioners of what might be called broadly the "progressivist" stripe, has meant that dialogue is the topic of the day and that promoting dialogue and the conditions which can support it are taken as central educational tasks. But the critics of dialogue raise issues that cannot easily be swept aside; and, in my view, some of these criticisms have raised deep problems for that approach.

Dialogue, Diversity, and Difference

As noted, one major point of criticism that has been raised against some accounts of dialogue is whether it is sufficiently sensitive to conditions of diversity, that is, the different forms of cultural communication, the different aims and values held by members of different groups, and the serious conflicts and histories of oppression and harm that have excluded marginalized groups from public and educational conversations in the past. Certainly, many accounts of dialogue have tried to respond positively to such criticisms. Yet even these attempts to respond have pointed up serious limitations in the standard accounts of dialogue.

What seems to recommend dialogue as a pedagogy is its capacity for active participation by all parties; its room for the co-construction of understanding or knowledge that can be negotiated between the perspectives of different members; its critical potential, which allows for questioning not only "within" the dialogue, but questioning its very terms and assumptions; and its open-endedness, its capacity for continuing and expanding the conversation to include multiple voices and perspectives — indeed, for many purposes, actively seeking them out. These are not trivial advantages, especially compared against many of the pedagogical practices currently in favor in education at all levels of schooling. Yet, as noted earlier, there is something self-confirming about this model: that its capacity to be self-corrective, in certain instances, is taken as proof that there is no legitimate "outside" to its procedures. One could call this "the hegemony of reasonableness": that precisely because dialogue seems to hold out the hand of inclusiveness and respect for all points of view, it makes those suspicious of its tacit rules of engagement, its "modes of address," as Ellsworth calls them, its scope of what is and is not up for discussion, appear as if they are at fault for remaining outside the conversation.

Dialogue runs up against difficulty in encounters with diversity. Not everyone speaks the same language; whose language will be used? Are the ground rules for participation, however thinly procedural they might appear, actually substantive restrictions on what can be talked about, on how things can be talked about, and so upon who can or will be part of the conversation? What are the limits of reflexivity within dialogue? Is the invitation to participate already a kind of co-optation of radical critique and rejectionism? Are the dialogical aims of consensus, provisional agreement, and even understanding (across unresolved differences) based upon ideals of harmony and community that are always on somebody’s terms and so threaten the maintenance of separate, self-determined identities? Finally, are there some differences that are simply unbridgeable in dialogue, gaps of understanding or belief that cannot be bridged — but which, in the attempt to bridge them, put some people more at risk more than others?

There are three broad ways that different models of dialogue have tried to address such issues of diversity. To an extent these cut across the six traditions described above, although some are more amenable to certain approaches than others. None of them adequately address, in my view, the criticisms of radical diversity just discussed. The first, pluralism, or the "melting pot" ideal, regards social and cultural diversity as a positive resource for the exchange of beliefs, values, and experiences that can inform and invigorate dialogue, but with the specific aims of reconciling these differences in agreements or compromises that combine the best elements from each perspective or form new, common understandings with which all parties can identify. In many instances, however, pluralism in this sense simply comes to the end of assimilating diverse groups into predominantly mainstream beliefs and values (though to a much lesser extent dominant or mainstream views may be modified over time as well). This asymmetry of change threatens, in the long run, to erase significant cultural difference, or to relegate it entirely to the private, not public sphere. The second approach to diversity, multiculturalism, perhaps the most widely held view in education today, emphasizes respecting (or celebrating) differences, not for the sake of assimilating them into dominant cultural forms, but to preserve them, both out of respect for the integrity of diverse cultural traditions and out of an appreciation of cultural variety for its own sake. However, this inclusive or celebratory attitude can also have the effect of exoticizing differences, rendering them quaint or interesting as artifacts, and not as critical points of reference against which to view one’s self. The framework within which multiculturalism often takes shape, a broad (and sometimes patronizing) "tolerance" for difference, leaves dominant beliefs and values largely unquestioned — indeed even insulated from challenge and change — because they are shielded within the comforting self-conception of openness and inclusivity. But as Cameron McCarthy has noted, multiculturalism means little if it is only other-regarding, and does not become an occasion for questioning the dominant cultural orientation as itself one of many, unprivileged, and just as quaint or strange (needing to be "tolerated") when viewed from the outside. Where cultural dominance comes from, and how it settles into a taken-for-grantedness that makes its own specificity invisible, is a question rarely explored within the multicultural framework. A third view, cosmopolitanism, of growing interest recently, emphasizes the unreconciled coexistence of diverse cultures and groups. Informed to some extent by a global perspective that recognizes not only the radical diversity of cultural difference, but also the attenuated circumstances that bring these cultures in contact with one another, this view of diversity (while often sharing many features with multiculturalism) acknowledges the limits of assimilation, agreement, or even understanding across certain cultural divides, and concludes that in many cases there must simply be an end to talk that seeks to bridge or minimize differences. Where such conversations exacerbate or heighten the awareness of disagreements or conflicting interests, continuing them may weaken rather than strengthen the prospects for a minimally harmonious condition in which each agrees to allow (and not necessarily respect or approve) the cultural domain and prerogatives of the other. The problem with this view, however, is that it abrogates — and sometimes prejudges and rejects out of hand — the value of engagement, excluding both the possibility of mutual accommodation or the possibility of a critical questioning of one view from a radically different other.

Because of the currency of debates over these three views, the question of the possibility and prospects for dialogue in contexts of diversity has become one of the central, if not the central, issues in contemporary educational theory and practice (to say nothing of larger social and political debates). The conventional view is that such dialogue is always a worthy effort and learning opportunity, even if in some cases it may unfortunately fall short of its ideals.

The problem here is that dialogue is variously viewed from these positions as a means of bridging differences, reconciling differences, coordinating action despite differences, or achieving understanding, respect, or tolerance in the face of differences. These objectives are clearly desirable under many circumstances, including educational circumstances; but these dynamics cannot be viewed symmetrically from all points of view. While some may view dialogue as a benefit, or a potential benefit, others may regard it as a threat, and others as an impossibility. The rejection of dialogue, or the refusal to submit one’s views to questioning, compromise, or renegotiation, is not always a mark of irrationality. The very aim of dialogue to speak and understand across differences is not an unalloyed benefit to all potential parties to such dialogue. Moreover, "difference" here is constituted as a dimension of diversity — categorical differentiation according to demographic, cultural, or identity categories. Dialogue tends to construct differences as instances of diverse values and points of view along continuums where middle grounds may exist, where commonalties may be found, or where translations across gulfs of misunderstanding may be achieved. Sometimes these are realistic prospects. Sometimes they are not; and where they are not, the reasonable gestures of inclusion made within dialogue can actually constitute co-opting or even coercive moves that put upon those with strong differences the burden of justifying why they will not participate.

In some of my earlier work, I suggested that dialogue could yield a range of outcomes, ranging from agreement; to a consensus (or in Rawlsian terms an "overlapping consensus") that falls short of full agreement; to an understanding that falls short of agreement or consensus; to a respectful tolerance that falls short of full understanding. Each of these, I suggested, can have fruitful educational benefits. My main point was that dialogue does not have to achieve agreement, consensus, or even understanding to be educationally (or socially and politically) worthwhile. Theoretically, this represented my departure from Habermasian or Gadamerian views of dialogue. I now think that this view suffered from three serious limitations. One is that these outcomes cannot be placed easily along a single continuum, like railroad stations at which a train may stop; that they are actually quite discontinuous sorts of paths, often entailing very different sorts of dialogical interactions — and so one cannot simply say that dialogue moves along a single "track" and gets as far along it as possible, aiming toward agreement or consensus, perhaps, but being satisfied with something less than that if full "success" cannot be achieved. Instead, assumptions at the start concerning which of several ends is possible or desirable have a determinative effect upon the form and tone of the type of dialogue in which one is engaged (and into which one is inviting others). This determination ("what type of dialogue are we having?") is one in which unilateral judgment and cultural dominance often play a central role.

The second failing was to underestimate the role of misunderstanding, and even incommensurability, as potentially necessary and even educationally beneficial ends under certain circumstances. I regarded them as failures of dialogue, or a sign that dialogue had not proceeded long enough. I stressed that one should never presume the outcome of incommensurability and suggested that one should always approach dialogue "as if" it need not end up that way. This was a mistake. There are instances in which the very encounter with a radically different, unreconciled, and unreconcilable point of view, value, voice, or belief can serve important educational purposes: to cause us to question the horizons of our own assumptions, to explore within ourselves (and not only within the other) the causes of why dialogue "fails," and to consider the possibility of a radically different way of approaching the world. Dialogue in the mode of resolving or dissolving differences provides no tools for coping with such encounters, or deriving meaning from them; it regards them as failures or breakdowns, and not as limitations within the model of dialogue itself.

The third failing was to conceive difference solely in the sense of categorical diversity. As Homi Bhabha and others argue, cultural difference can be taken in a different way: as a less stable, noncategorical dimension that is a feature of lived experience and identity. From this standpoint, differences are enacted. They change over time. They take shape differently in varied contexts. They surpass our attempts to classify or define them. Ellsworth puts it well, that the purpose of dialogue is not just speaking across given positions of difference, but a relation in which those very positions can be (need to be) questioned. Difference, then, is more than a matter of multicultural diversity, of speaking within and across stable identities; it is a challenge to these in three ways, which I have sketched in more recent work as differences within, differences beyond, and differences against. Respectively, these three phrases refer to the ways in which: (1) difference stands not only as an external feature of the "other," but as an unexplored and unrecognized dimension of one’s self (for example, in the ways by which heterosexuality is defined and defended implicitly as not-homosexual, thereby invoking its "opposite" as a part of its own self-conception); (2) difference exceeds categories of understanding, challenging these in ways that confound conventional vocabularies and assumptions (for example, when racial categories such as "black" and "white" become denaturalized and subject to all sorts of redefinitions, including those of skin color themselves (no one actually has black or white skin), the conflating of racial with national or ethnic differences, the emphasis on hybrid, creole, or border identities, and so on); and (3) difference is defined by its resistance, defined against dominant norms, and its persistent refusal to allow itself to be characterized from dominant, conventional points of view. In each of these three ways difference poses a fundamental challenge to views of dialogue oriented around achieving understanding or agreement — each, in its own way, is a repudiation of convergent models of discourse generally, and each, in its own way, resists the categorical characterization of diversity — no category can possibly contain these sorts of difference.

It is possible to put the point even more strongly: that the effect of traditional views of dialogue has been to "domesticate" difference: to make it safe and comprehensible by regarding all differences as elements of mere variation (diversity), and hence as starting points of potential reconciliation. This is not a neutral standpoint, even as it represents itself as such; it misses deeper, more radical conceptions of difference.

Dialogue as Decontextualized Pedagogy

The crucial shift in perspective outlined here is from a prescriptive model of dialogue as a neutral communicative process, a procedure in which all participants are treated equally, concerned only with the search for knowledge, understanding, and perhaps agreement, to dialogue as a situated practice, one implicated by the particulars of who, when, where, and how the dialogue takes place. The elevation of dialogue as a general pedagogical method abstracts its operations from those particulars and, as noted earlier, treats deviations from that ideal as either illegitimate violations of its rules, or as unfortunate shortcomings that can be remedied through the application of more of the same — continuing with dialogue until these failures of understanding or agreement can be remedied. Radical difference, difference that resists accommodation or assimilation, is rendered inexplicable or perverse. But when one examines the who, when, where, and how of dialogue, such characterizations become much more difficult to defend.

Who. The first issue begins with the growing diversity of classrooms (at all levels of education) and the increasing awareness of the margins or borders of common school culture as it interacts with the very different values and orientations that students bring to the classroom. The conditions of globalization and mobility have promoted both direct forms of migration across national/cultural categories and (especially with the rise of new communication and information technologies) an increasing proximity and interaction of multiple lines of national/cultural influence. In this context, the central assumptions of common schooling — of a canon of texts, of a shared historical tradition, of a common language — are thrown into question, since even where such elements might be defended, their value and significance are going to be regarded differently from different positions as teachers and students. In some cases they will be directly challenged. The shift to a dialogical approach, in itself, does not remedy these conflicts; and when more radical conceptions of difference are at stake, the very notion of "remedying" such conflicts and disagreements becomes deeply problematic.

A dialogue is not an engagement of two (or more) abstract persons, but of people with characteristics, styles, values, and assumptions that shape the particular ways in which they engage in discourse. Any prescriptive conception of dialogue must confront the challenge of acknowledging persons who do not engage in communication through those forms, and who might in fact be excluded or disadvantaged by them. Conversely, an account of dialogue that acknowledges the enormous multiplicity of forms in which people from different cultures do enact pedagogical communicative relations (let alone communicative relations generally) needs to address the question of why some versions are rewarded with the prescriptive label "dialogue" and others not.

Aside from the multiplicity of communicative forms, there is also a multiplicity of communicative purposes in dialogue. In many contexts, for example, the formation and negotiation of identity may constitute the primary purpose in mind for some participants in a dialogical relation, supplanting more overt teaching-learning goals. Such dynamics may be only partly intended or conscious (and hence only partly susceptible to reflection or change). Participation in dialogue, even at the micro-level of apparent personal "choice," is not simply a matter of choice. The utterances that comprise an ongoing dialogue are already made (or not made) in the context of an awareness of the reactions — real, anticipated, or imagined — of other participants. The more that one pushes this sort of analysis, the more the achievement, or suppression, of dialogical possibilities comes to be seen as an expression of a group interdynamic, and not something resulting simply from the choices and actions of individuals.

When. We do not just use language; language uses us. As Bakhtin argued, the nature of discourse is that the language we encounter already has a history; the words that we speak have been spoken by others before us (he calls this "the internal dialogism of the word"). As a result, what we speak always means more than we mean to say; the language that we use carries with it implications, connotations, and consequences that we can only partly intend. The words that others hear from us, how they understand them, and what they say in response are beyond our unilateral control. The multivalence of discourse situates specific speech acts or relations in a web of potential significations that is indeterminate, nonlinear, and highly susceptible to the effects of context and cultural difference.

A dialogue is not simply a momentary engagement between two or more people; it is a discursive relation situated against the background of previous relations involving them and the relation of what they are speaking today to the history of those words spoken before them. These background conditions are also not simply matters of choice, and they impinge upon the dialogical relation in ways that may shape or limit the possibilities of communication and understanding. Often these relations are expressed as forms of power or privilege because the relative positions of people place asymmetrical constraints on who can speak, who can be heard, and who has a stake in maintaining a particular dialogue, or in challenging it. The prescriptive model of dialogue has reinforced a view of dialogue as a finite and bounded engagement, often described with little or no context, and with scant consideration given to what might have transpired before or may transpire after the dialogue at hand. This has tended to support the idea of a dialogue as a unitary, goal-oriented conversation with a discrete purpose and a beginning, middle, and end, not as a slice of an ongoing communicative relation, as it nearly always is in educational settings.

Where. Recent years have seen a growth of interest in such problems as situated cognition, group learning, the relation of expert and novice understandings, real-world problem solving, distributed intelligence, and a whole range of similar notions that address in different ways the actual means by which the learning of individuals occurs in the contexts of existing social relations and practice.

These concerns apply directly to the matter of dialogue. The situatedness of dialogue, considered as a discursive practice, means that the dialogical relation depends not only upon what people are saying to each other, but the context in which they come together (the classroom or the cafeteria, for example), where they are positioned in relation to each other (standing, sitting, or communicating on-line), and what other gestures or activities work with or against the grain of the interaction. Dialogue has a materiality, which means paying attention to both facilitating and inhibitive characteristics in the circumstances under which it takes place. Nor is it simply a matter of the present context at hand, but also other contexts — including anticipated future contexts of need or use — that shape the understanding of purposes that guide or direct a dialogical engagement. For example, interactions at home, in the playground or lunchroom, or on the street before or after school may constitute contexts of teaching and learning that are at least as important for certain participants as the interaction in the classroom; and relative importance aside, they certainly impinge upon the thoughts, feelings, and motivations participants bring to the classroom.

How. Another aspect of this situatedness, or materiality, is that the texts and objects of representation that mediate classroom discourse can have distinctive effects on what can be said and how it can be understood. Where interaction takes place in an immediate, face-to-face circumstance, these "texts" include not only the words themselves, but facial expressions, gestures, and similar representational forms. Yet dialogue often also takes place in mediated forms: a dialogue between a book’s reader and its author; a dialogue between correspondents writing to one another; a dialogue over a telephone; a dialogue over e-mail; and so on. The tendency of previous accounts of dialogue has been to ignore such factors or, if they are considered at all, to relegate them to trivial significance compared to what the words themselves express. Yet substantial research across a range of fields has highlighted the ways in which the circumstances of form and medium are not trivial, but can influence what is said and how it is understood, and the ways in which these media are signifying elements themselves.

In these four ways, then, the prescriptive account of dialogue has been impeded by the formal, idealized models through which it has been characterized: impeded because these models have often not taken account of the situated, relational, material circumstances in which such discursive practices actually take place. Attending to the social dynamics and contexts of classroom discourse heightens the awareness of the complexities and difficulties of changing specific elements within larger communities of practice. These communities may be the primary shapers of teaching and learning processes, but not always in ways that serve intended or ideal educational objectives; other purposes, such as identity formation or negotiating relations of group solidarity, may predominate. The power of such social processes may restrict lines of inquiry, distort dialogical interactions, and silence perspectives in ways that conflict with the explicit purposes of education.

Rethinking dialogue along these lines holds promise for developing theoretical accounts of dialogue that are richer, more complex, and better attuned to the material circumstances of pedagogical practice. Dialogue, from this standpoint, cannot be viewed simply as a form of question and answer, but as a relation constituted in a web of relations among multiple forms of communication, human practices, and mediating objects or texts.

Criticizing the Decontextualized, Prescriptive Model of Dialogue

The major contemporary critic of the prescriptive model of dialogue and its virtually unquestioned role in critical pedagogy is Elizabeth Ellsworth. Her current critique focuses on considering dialogue as a "mode of address," one that positions teacher and learner in a determinate relation (even one that is ostensibly egalitarian) and, in so doing, constrains the possibilities of communicative exchange, no matter how "open" it aspires to be. Instead, Ellsworth calls for "pedagogical modes of address that aren’t founded on striving for and desiring certainty, continuity, and control" and "pedagogical modes of address that multiply and set in motion the positions from which they can be ‘met’ and responded to." Referring in part directly to some of my own earlier work, she writes,

By communicative dialogue, I mean a controlled process of interaction that seeks successful communication, defined as the moment of full understanding. For those who advocate it in education, communicative dialogue drives toward mutual understanding as a pedagogical ideal....In other words, what must come first in communicative dialogue is understanding — that is, a supposedly innocent, disinterested reading of the other’s message. Then disagreement is allowed....What communicative dialogue cannot tolerate, what it must exclude, is the one who says, ‘Our differences are such that you cannot understand me, and I cannot understand you.’....Communicative dialogue works only when we act as if its mode of address is a neutral conduit of reality, and not itself a rhetoric — not itself a mediation of knowledge and of its participants relations to knowledge.

I think that this criticism is basically correct: a conception of dialogue based on the idea that "successful communication" can only mean "full understanding," and the idea that dialogue is, or can be, a "neutral conduit of reality," itself proof from question, is entirely inadequate — even damaging. There are many cases in which the striving for understanding (or agreement) at all costs will run roughshod over individual or group differences that cannot be bridged easily or reconciled with dominant understandings. It must be seen that dialogue can be "successful" just in the sense of bringing to light the experience and perspectives of others quite different from ourselves (and this can be a kind of success even when we cannot entirely understand, let along agree, with them). Ellsworth is right that the ideal of "dialogue" can become an actual impediment to human freedom, diversity, and co-existence. Moreover, Ellsworth is also right that, if the implicit communicative rules and aims of a dialogical engagement cannot themselves be questioned or challenged, reflexively, from within the dialogue, then not only will certain voices or perspectives be excluded from possible participation, but the medium of dialogue itself becomes a way of structuring interpersonal knowledge and understanding, in a decidedly non-neutral way, without recourse to considering alternative frames that might be possible.

What puzzles me about Ellsworth’s criticisms of "communicative dialogue" is, first of all, to wonder where she finds such a caricatured view of dialogue in my own work (where I have repeatedly said that knowledge, agreement, and understanding are only some of the potential outcomes of dialogue; that dialogue sometimes encounters differences that surpass our ability to understand them and lead to unreconciled disagreement; and that these, too, can foster important educational benefits and learning opportunities). But of greater concern to me is whether Ellsworth thinks that having disposed of "communicative dialogue," in the sense she describes it, one has refuted somehow the idea of dialogue itself. Sometimes she has written as if she thinks that she has. But in her latest work, in fact, she actually defends an alternative ideal of dialogue, which she terms (following psychoanalytic theory) "analytic dialogue":

What gets "analyzed" the route of a reading. How did you arrive at this interpretation, without knowing it — maybe even without desiring it? How have your/our passages through history, power, desire, and language on the way to this interpretation become integral parts of the very structure of the interpretation — of our knowledge?

I believe that Ellsworth is exploring here a crucial sense in which any communicative form, including "dialogue," needs to be subject to question itself. No medium is neutral, no utterance or observation can claim an entirely disinterested or non-positioned vantage point. Whenever any pedagogical practice or relation becomes "naturalized" and comes to be seen as the only possibility, the best possibility, or the most "politically correct" possibility, it becomes (ironically) an impediment to human freedom, diversity, exploration, and — therefore — the possibilities of learning and discovery. As I have noted, in many accounts of dialogue and pedagogy the "fetishization" of dialogue has obscured some of its real limitations and contradictions. Moreover, the proclamation of any particular dialogical genre as the instrument of human emancipation (such as the Socratic method, Freirean pedagogy, or a Habermasian search for consensus) will inevitably exclude, silence, or normalize others from radically different subject positions. I and other theorists working on these topics owe appreciation to Ellsworth, Patti Lather, Mary Leach, Alison Jones, and other feminist poststructural critics for pressing this issue so strongly.

Engaging the Criticisms: From Prescriptivism to the Practice of Dialogue

I would like to think that I am open to criticism and try to learn from my mistakes. Still, I persist in thinking that some of these very criticisms reinforce the value of "dialogue" in some sense, if in a very different sense from its conventional uses (and, as I have noted, even Ellsworth wants to defend a conception of "dialogue"). I believe that this alternative view of dialogue begins by questioning two elements in most conventional views of dialogue: prescriptivism and proceduralism. Questioning prescriptivism entails reflecting on the ways in which "dialogue" has become a kind of unquestioned ideal, a norm, and a rhetorical device. To invite others into dialogue is seen as an unassailable gesture of good will. Who could criticize or reject such a gesture, except the ill-willed, the alienated, the recalcitrant? Such a stance, however, ignores the many ways in which this invitation may not be open-ended or neutral, or not experienced as such by others, even when it is intended to be. To enter into a conversation is to accept a set of tacit communicative norms; it is to run the (often asymmetrical) risks of disclosure; it is to undertake to explain one’s self, perhaps justify one’s self, under questioning; it is to submit to a set of assumptions about what "the subject" of the dialogue is about, and what it is not. The point here is not that these commitments are never fair expectations to have of participants to a dialogue; it is to acknowledge that for many parties, under specific circumstances, they represent a kind of entrapment, a kind of co-optation, in which some persons have more to lose than do others. From this standpoint the humanistic ideal of "engaging in dialogue" comes to be seen as subtly coercive, even threatening. And for theorists, such as myself, who have tended to favor persistence and "keeping the dialogue going" as prescriptive norms, this criticism provides a much-needed rebuke.

Alison Jones provides a fascinating analysis of this problem in practice. Juxtaposing the ideal versions of dialogical pedagogy with the realities of conversation in a class where she brings together Maori (native, minority culture) and Pakeha (European, dominant culture) students in New Zealand, she notes that "an ideal dialogical model for the classroom asserts that stories and meanings of less powerful as well as more powerful groups will intermingle and ‘be heard’ in mutual communication and progressive understanding....[It] assumes that the opportunity for subordinate groups to express themselves in the critical classroom becomes an opportunity for ‘empowerment.’" But Jones shows how in practice even this apparently benevolent, receptive stance by those in relative positions of power can in fact reinscribe their privileges and advantages: "Border-crossing and ‘recognitions of difference’ turns out to be access for dominant groups to the thoughts, cultures, lives of others." She even terms this a "cannibal desire to ‘know the other,’" a further sort of exploitation, which the Maori students understandably resist:

In the midst of all this mess and discomfort, I wonder what is the pleasure in (ethnic) difference for the dominant group in education? Why do we repeat the phrase in our theories and writing, at this fashionable moment of respect for, if not celebration of, difference? Apart from a certain voyeurism, is it not that "we" (the liberal/radical dominant) can be the other-who-now-speaks that we are part of the scene of redemption; that we are not the unfashionable colonizer/oppressor whose despised description fills our textbooks, and from whom we...are usually pleasingly distanced?...[We] seek liberation, through ‘your’ dialogue with us. Touched by your attention, we are included with you, and therefore cleaned from the taint of colonization and power which excludes.

The second, related issue is the proceduralism of most accounts of dialogue: the characterization of dialogue in terms of a particular set of communicative norms, and the response, when conflict or friction arises, that the resolution of these can (and should) take place through a reinvigorated application of those same norms. It is clear from Ellsworth’s critique that this shields from questioning or criticism those norms themselves. Yet here we encounter a paradox (the first of several to follow); one that begins to turn the discussion of dialogue into a different, more productive theoretical vein. For if questioning the restrictive norms of dialogue is regarded as a good thing, is it not at least in part so that a fuller, fairer, more inclusive dialogue might be made possible? If persons choose to withdraw from a dialogue with those who do not or cannot understand them, is it not in part so that they are able to enter a dialogue with others who can understand them? If "analytic dialogue" seeks (rightly, I would say) to uncover the nonneutral, historically specific conditions under which its own interpretations proceed, is this not so that others might come to share the same understanding, at least in part, about these conditions? It seems strange indeed to imagine a dialogue in which every understanding emerges as entirely idiosyncratic and separate from every other; or one that is so persistently pulling up the roots of its own genealogy that the participants never talk about the topic at hand.

One of the most admirable elements in Ellsworth’s book is her honesty about some of the paradoxes of her own pedagogical practice. She writes, "At the same time, as an educator, I can’t pretend that my own teaching practices haven’t been troubled by the paradoxes and impossibilities of communicative dialogue, of democracy, and of teaching itself." The tone of confession in such passages is striking, as if she is trying to reform, but keeps backsliding into disreputable misconduct. I would want to reframe the issue in a different way: Our teaching practices remain troubled by the paradoxes and impossibilities of "communicative dialogue" because there is no way to engage in teaching without encountering them. A fuller, less dyadic, understanding of dialogue must recognize multiple moments within it, some inevitably "communicative," others perhaps "analytic" (in Ellsworth’s senses); some convergent toward agreement and understanding, others transgressive and dispersive (Bakhtin referees to these as the centrifugal and centripetal forces within any dialogical engagement); some "friendly" and others "disputatious"; and so on. Indeed, once one starts thinking of dialogue in such terms, the more difficult it is to maintain the dyadic character of these either/ors. For in any ongoing dialogue, all of these moments may recur, with no one of them defining "dialogue" as such. Such a view of diverse forms, purposes, and relations is partly a corrective to what I have called the "fetishization" of dialogue, or the reification of any particular form (even including the "analytic").

In short, the criticisms posed against dialogue by Ellsworth and others have had a tremendously constructive benefit in unsettling the prescriptive account that has predominated in educational discussions. Her challenges to the silences, exclusions, and coercive or co-opting elements in dialogue, which challenge its self-conception as something open, neutral, and inviting to all, need to be addressed directly. What these criticisms have done is to refocus attention on the practice of dialogue, with its tensions, paradoxes, and material effects on those who are not willing or able to participate in educational discussions in that manner. Engaging in this practice requires awareness of these difficulties and dilemmas, and an acknowledgement that particular forms of dialogue cannot serve the very aims that they avow.

Yet these criticisms, in turn, confront some of their own difficulties. The first of these is that the elevation of difference, while an invaluable corrective to those views of "communicative dialogue" that emphasize the pursuit of agreement, consensus, or understanding as the only legitimate outcomes of dialogue, cannot stand as an absolute principle in its stead. Agreement, consensus, or understanding (which, as I have discussed, are very different sorts of outcomes) may sometimes be unobtainable in dialogue, and — even where attainable — they may be problematic, provisional, and properly subject to questions concerning how and one whose terms they have been obtained. Fair enough. But Ellsworth often writes as if these were inherently undesirable outcomes, never justifiable as voluntary and intersubjective. This cannot be true, both as a matter of experience and of history, where such outcomes — even in the face of deep difference and conflict — have been satisfactorily arrived at; and as a matter of social and political principle, where there are occasions in which the pursuit of such outcomes, with all their risks of difficulty and failure, are the sole alternative to violent adjudications of conflict. What the theory of dialogue needs is a modulated account of where and when such outcomes can be secured, and how to be suspicious of them while also recognizing their value for different groups’ purposes. If asymmetrical and unequal power were always conditions that abrogated the value of human engagement (including communicative engagement), then there would be no legitimate engagements at all, because power is never entirely asymmetrical and unequal. And, as I tried to show earlier, alternatives such as "analytic dialogue" believe in the value of understanding and agreement too.

Second, and building upon this point, the corrective elevation of radical difference sometimes segues into the presumption of incommensurability. I have addressed this issue in other writings. In the face of radical difference, misunderstanding or nonunderstanding are certainly possibilities. Sometimes dialogue reaches an impasse. But this account, taken on its own, is an oversimplification. For one thing, misunderstanding is not an all-or-nothing state; in real, situated contexts, degrees of misunderstanding are mixed with degrees of understanding, and the practical question at hand is where and for what purposes (and for whom) the degree of understanding is sufficient for the purposes — including the educational purposes — at hand. Too much rhetorical ink has been spilled, in my view, drawing the false alternatives of a realist, objectivist view of dialogue centered on a single "Truth," and a radically incommensurable alternative in which all knowledge is politically contested and culturally idiosyncratic. We need to get beyond these useless alternatives, especially if we are to speak in any constructive way about educational interactions. The paradoxical challenge here is to recognize the excess of meaning, the differend, as Lyotard calls it, that may be beyond translation or comprehension in many, even most, communicative encounters, and to realize that sometimes this excess may be of crucial import in adjudicating, or failing to adjudicate, a serious difference of belief or value; while at the same time recognizing the practical need to pursue the degrees of understanding appropriate to particular purposes, including educational purposes. Sometimes, indeed, this endeavor fails — and, as noted earlier, this failure can have crucial educational import in alerting us to the horizons of our own assumptions, to our own culpability in why the dialogue "failed," and to the possibility of considering a radically different way of approaching the world. But if one believed truly that such encounters always failed, it is unclear what meaning "education" could ever have.

Third, and finally, the juxtaposition of what I have called the prescriptive perspective (represented by formal, idealized models), and the practical perspective (represented by situated, politically critical analysis) on dialogue, itself draws an overly sharp distinction. For reasons that cannot be developed fully here, social practices always entail at least implicitly prescriptive norms, and in this sense always run the risk of being impositional, normalizing, and exploitative of relatively powerless persons or groups. Alternative social practices may avoid those failings, but replace them with others. Critical dialogue, communicative dialogue, analytical dialogue, and every other educational approach entail their own latent prescriptions — even apart from those they try to make explicit and open to question — and so inevitably encounter a limit to their capacities to be self-reflexive and self-problematizing. Some communicative relations, such as Habermas’s acceptance of the legitimacy of "validity claims" and challenges within a discursive context, or Ellsworth’s advocacy of "modes of address that multiply and set in motion the positions from which they can be ‘met’ and responded to," bring these possibilities of reflexivity more to the surface; other communicative relations tend to be more oblivious or even resistant to such reflexivity. But paradoxically, again, it may actually be that those very communicative relations that try to be most open about their implicit commitments and prescriptions may be for that very reason more difficult to diagnose in terms of their blind spots and, hence, more difficult to resist. Or, to put this a different way, those modes of dialogue that put the greatest emphasis on criticality and inclusivity may also be the most subtly co-opting and normalizing. Such a recognition unsettles critical pedagogies of all sorts, whether feminist or Freirean, rationalist or deconstructionist.