The complexity revolution that emerged from the work of Lorenz (Gleick, 1988), Prigogine (Prigogine & Stengers, 1984), and many others has certainly spread into management and organizational theory, with implications for economics (Arthur, 1994), human development (Senge, 1990), and organizational theory (Wheatley, 1992). Communication scholars have also recognized a tremendous explanatory potential in complexity (Anderson & Houston, 1997; Hammond, 1997; Hawes, 1999), while noting its limitations (Woods & Simpson, 2002).

With this increased scholarship has come greater debate on how complexity and self-organization work within social situations. As Woods and Simpson (2002) have asserted, creating working explanations is problematic. Some frustrated social scientists (Letiche, 2000: 58) have called the complexity paradigm “messy.” Axelrod and Cohen (1999: 15) say, “there is little convergence among theorists who have begun to study complex systems as a class. It is not a field in which a crisp and unified theory has already been developed, nor is one expected in the next few years.” In addition to concerns over the instrumentality of complexity theory, as the paradigm matures and penetrates deeper into each discipline the conversation between disciplines seems to diminish. As a result, important conceptual breakthroughs in one area go unnoticed by another.

This article makes two important contributions. The first is to build a conceptual two-way bridge between those who are concerned with communication in scientific disciplines and those who are concerned with communication from social scientific paradigms. Building this conceptual bridge leads to an important redefinition of communication focusing on dialogic processes that lead to social self-organization rather than linguistic processes that lead to a rhetorical destination. Implied in this redefinition is a position that self-organizing systems must define themselves, describe themselves, mark the common ground of order and disorder, and then redefine themselves. Our redefinition toward a process orientation of communication lays the foundation for the second section where we examine the implications of this new definition and identify five different “attractors” around which communicative selforganization occurs. Thus, we show how complexity “works” through dialogic means to create social self-organization in organizational environments, extending our understanding of dialogue “as a way of interacting that facilitates the construction of meaning” (De Weerdt, 1999).


We can redefine communication from two directions. The first is from the scientific complexity paradigm, where we will show how those who describe self-organizing processes redefine communication by arguing that it is a nonreversible and nonpredictive process that is essential for self-organization, defining and confining the social organization to certain structurating possibilities. The second direction is from communication theory, where we will note how some communication scholars have redefined communication in ways that align with scientific complexity concerns. Both the scientific and communication approaches to complexity point toward a dialogic view of communication that is best explained by the principles of self-organization.

Of course, the very term “self-organization” has multiple meanings. But communication defined in the self-organizing frame suggests that a system is in an ongoing dialogue with itself in order to define itself, describe itself, mark common sites of order and coherence, mark common sites of disorder and incoherence, and redefine itself.


We do not intend to provide an exhaustive literature review of the many contributions that scientific theories of complexity have made toward a redefinition of communication. We are also limited in our review of the scientific literature, focusing mainly on ideas accessible to social sciences. We do, however, want to highlight four critical points coming from complexity science that tear communication from its previous linear paradigm (Carey, 1989) and center its function within complex systems. Scientific theories have shown that communication is nonreversible, nonpredictive, fundamental for self-organization, and system defining.

Communication is non-reversible

Traditional, rhetorical views of communication popular through the nineteenth and twentieth centuries suggested that communication was a bipolar process that allowed one person to target another, convince the other of their rightness, and then enjoy the power of their persuasive role. If someone “changed their mind,” the rhetorical process was reversed and they were reconverted to the old position. Entire institutions, parliaments, and legal systems were founded on this linear notion that is based on a distinguishable dualism between self and other.

In science, Prigogine (Prigogine & Stengers, 1984) changed that notion, suggesting that time (and rhetorical positions) cannot be reversed. According to Prigogine, in complex systems we cannot go back, only forward. All things (including speaking) cannot be reversed. Prigogine and Stengers (1984: 12) say:

During the nineteenth century, the final state of thermodynamic evolution was at the center of scientific research. This was equilibrium thermodynamics. Irreversible processes were looked down on as nuisances, as disturbances, as subjects not worthy of study. Today, the situation has completely changed. We now know that far from equilibrium, new types of structures may originate spontaneously.

All communication has an element of the spontaneous. Even a carefully scripted television program can have spontaneous meaning as it is watched in different social and cultural contexts by viewers with different values. Whatever meanings are made go forward and cannot be erased.

Communication is non-predictive

The spontaneity described by Prigogine comes about as a result of the inherent randomness in any complex system. It is the tension between disorder created by randomness and order imposed by shared meaning and experience that creates the need to communicate. Crutchfield (1986: 46) said that the idea of self-organization within the complexity paradigm “implies new fundamental limits on the ability to make predictions. On the other hand, the determinism inherent in chaos implies that random phenomena are more predictable than had been thought.” In other words, we communicate to find meaning and create order in an equivocal world, but our communication processes create disorder at the same time. Cohen (2002) describes a similar tension in biological systems.

Communication is fundamental to self-organization

Cramer (1993) argues that self-organizing communication is not a property merely of living structures but of all structures. He says that the randomness brings limited agency to all structures and that “true selforganizing is a property of the entire system” (Cramer, 1993: 171). Cramer’s argument is holistic and implies a material realization that a dialogic information exchange is a central part of self-organization, and that the only real dualisms are symbolically created.

The assertion that self-organization is a fundamental property of matter means, at the same time, that this is a priori filled with ideas. It carries within it the idea of its self-organization, its self-realization, all its blueprints and physical manifestations. Accordingly, the idea of human consciousness must have existed as a possibility at the very moment of the big bang. From this point of view, there is no opposition between spirit and matter. In any case, spirit cannot have arisen as a super-structure from matter. The opposite is more likely; a matter devoid of ideas, without the idea of self-organization, does not exist, any more than weightless matter exists. (Cramer, 1993: 172)

Communication is life giving and system building

Cramer’s notion that consciousness and self-organization (and thus communication) are essentially life giving is shared by other scientists. Complexity biologists like Varela (1987) have suggested that information is life giving and that the movement of information, or communication, is essentially life itself. Information not only informs the other, but it creates the self. Self-reference, according to complexity biologists, is the essential life-giving outcome of communication. This moves biology to a broader and more holistic view of life, articulated by Botkin (1990: 7):

We are accustomed to thinking of life as a characteristic of individual organisms. Individuals are alive, but an individual cannot sustain life. Life is sustained only by a group of organisms of many species—not simply by a hoard or mob, but a certain kind of system composed of many individuals of different species—and their environment, making together a network of living and non-living parts that can maintain the flow of energy and the cycling of chemical elements that, in turn, support life.

Note that while many of these scholars infer new characteristics, they do not directly address communication. They describe the critical function of networks, information sharing, and talk about self-consciousness and self-reference. They describe the flow of energy and the flow of information. What Wheatley (1992: 105) described is certainly well within the domain of communication:

Information is unique as a resource because of its capacity to generate itself. It’s the solar energy of organization—inexhaustible, with new progeny emerging every time information meets with itself. As long as there are senders and receivers linked together in a context, fertility abounds. All that is needed is freedom of circulation to guarantee new births.


Simon (1962) is the first known reference in communication literature to self-organization. While many communication scholars argued for a more nonlinear approach drawing from cybernetics and second-order cybernetics (Krippendorff, 1994), it was not until the late 1980s when the ideas from science began making their way into mainstream communication theory. Communication scholars, friendly to the ideas of the new French philosophers (Deleuze & Guittari, 1987; Latour, 1988), first found the vocabulary of complexity helpful in rethinking communication (Anderson & Houston, 1997; Hammond, 1997). Later, some communication scholars went beyond drawing mere conceptual distinctions to using complexity in research designs and in making broad philosophical claims (Hawes, 1999).

We are now at a point where we can begin to see how the ideas of selforganization inform our understanding of our most obvious environment: our social interactions. The redefinition of communication in complexity terms features that abandonment of the traditional, rhetorical view of communication as persuasion. It also forces a repositioning of the sensemaking (Weick, 1995) paradigm, where communication reduces equivocality. The complexity paradigm in communication sees the communicative act as a source of both order and disorder present in all systems.

Indeed, Luhmann (1984) claims that communication is the medium of all systems. Taylor (2001: 149) explains that Luhmann argues that “the common tendency to see communication as a consequence of one person acting and another reacting is to invert the real relationship between action and communication. It is not action that produces communication. Only communication can produce action.” Luhmann (1992: 251) states, “only within such a network of communication is what we understand as action created.” Following Luhmann’s idea that action does not produce communication, Baraldi (1993) explains that in this perspective, communication must be seen as the process of creating continued communication. In addition, Hammond (1997), influenced by the scientific paradigm, claims that communication cannot be separated from the system itself. Thus, communication produces the interdependent interaction in a system, rather than the interdependent interactions producing certain types of communication. And consequently, communication continually sustains and reproduces the system. “Communication is self-regenerative. In other words, one thing leads to another, to produce sustained interaction” (Taylor, 2001: 149). In simple terms, communication theory is moving from a simplistic persuasive model based on a transmission metaphor, to a new set of framing metaphors and theories based directly or indirectly on complexity. Communication is no longer seen as a rhetorical destination but as a process shared within a social system.

Some scholars are beginning to recognize this need to understand communication as the basis of all systems, and specifically self-organizing systems, and to call for scholars to begin uncovering the specific communicative processes by which self-organization in complex systems occurs. Salazar (2002), in discussing self-organization and the generation of group creativity—creativity being a salient issue for complexity scholars, such as Coleman (1999)—claims that there is a need to identify the characteristics of group adaptation in fluctuating environments, “as well as identifying the initial conditions that give rise to system features that enable creative activity” (Salazar, 2002: 180). Dooley and Van de Ven (1999) point out the limitations to current thinking about the complexity perspective, arguing that there is too much emphasis on the implications of complexity and that scholars are overlooking the need to uncover the mechanisms that generate it.

Menz (1999: 107) explains, “If communication really is the medium of all systems (Luhmann, 1984), then self-organization must above all constitute itself by and be identifiable through specific processes and forms of communication.” He claims, “we need to inquire into the specifics of self-organizing communication and examine its defining characteristics.” And Salazar (2002) calls for communication scholars to actively become a central part of this defining process. Luhmann (1992: 251-2) explains that to begin this defining process, “one must not begin with the concept of action, but with the concept of communication.”

Making the claim that communication is the medium of all systems, especially social systems (Luhmann 1984, 1992), an analysis of some of the many definitions of complexity and self-organization will reveal their underlying communicative assumptions that support this claim. Although these definitions are common understanding among complexity scholars, reevaluating them from a communication perspective will illuminate the communicative nature of complexity and self-organization.


It is often difficult to see how self-organization “works” dialogically within social systems through the communication process because selforganizing occurs in real-time relationships that surface and submerge in our perceptions. Some work has already occurred in this area. According to Axelrod and Cohen (1999), three key terms describe the complex framework of self-organization in social systems: agent, strategy, and population. Agents use strategies to understand and exchange information with a larger population of agents. These agents adapt to their environment as it changes, select which information is necessary to interpret, and interact accordingly to try to bring order to their situations. In other words, Axelrod and Cohen say, people, embedded with identity, communicate with other members of their organization as they make meaning around new phenomena and hope for a specific outcome. They do this in order to keep up with constant challenges and continued change. These people communicate individually or collectively about those things they encounter in their organization that are pertinent to their surroundings, trying to keep up with the constant changes in the organization and the field of the larger economy that surrounds it.

An “economy,” as described by Axelrod and Cohen, is a dynamic, selforganizing system. According to Cramer (1993), there must be two elements. The first is a tension, difference, misunderstanding, or underdetermination where meaning is in dispute. It is only when we don’t understand one another that we find the need to communicate. The second is an attractor, domain, or field that binds our social system. An attractor is like a domain, constrained by an underdetermined, fixed point with definable boundaries but unlimited internal possibilities. For example, the weather has definable boundaries. The temperature will never reach 150 degrees Fahrenheit, nor will the wind blow at 200 miles per hour in any weather pattern, but within a weather pattern there are infinite possibilities for wind direction, duration, and other measures. There are enough patterns that prediction is a hope, though never a complete possibility.

Cohen and Stewart (1994) describe such an attractor, where they suggest that a realistic portrait of predator-prey dynamics must include both an attractor and a repeller. All populations live within the tension between these two boundaries. If the prey becomes too populous, it allows the predator population to grow, who in turn kill more prey. About these population dynamics they say, “Attractors are the things that the dynamics converge toward if you wait long enough; but once they reach the attractors, they diverge again” (Cohen & Stewart, 1994: 206).

From our observations of real-time dialogues and assessment of the literature, we believe that certain tensions are present in every dialogic encounter, and that they cluster around certain attractors. These form permanent questions that are always before us in dialogic situations. These questions may be subtitle or direct, clear or unclear, obvious or opaque, but they are ever present in social self-organization. In other words, they are always part of the dialogue.

It is clear that our list is not exhaustive, nor are our examples undisputed. These tensions, although not always obvious, are layered within the dialogue and subjectively experienced. They are “once-occurent events of Being” (Bakhtin, 1993; Shotter, 2000) that can be subtle, elusive, powerful, clear, disputed, disruptive, revealing, and even transformative. They are always in dispute and thus are the rhetorical artefacts of social self-organization. As participants engage one another in dialogic encounters in organization, they enter into a nonreversible field similar to that described by Prigogine (Prigogine & Stengers, 1984) or Hayles (1988) that opens possibilities for change in communicators, their identities, the group, the meaning, and so on.

It also opens the possibility of our own confusion. Shotter (2000), in a point very similar to those made by Prigogine and Hayles, cautions us about the problem with this kind of scholarship when he says that we are attempting to characterize a flow while we ourselves are in the current: “Our everyday ways of communicating and understanding each other from within this flow are not best understood by being viewed through the refined and systematized products that emerge from our more everyday, informal modes of communication” (Shotter, 2000: 121). He argues that it is extremely difficult to move “upstream” and make systematic observations about the “special moments” when we are in “an immediate living contact with each other” (Shotter, 2000: 121).

The task is indeed difficult, but if we are to adopt the complexity paradigm and use it to describe how this “works,” we must learn to live with this difficulty. We describe each of these observed tensions as just that: tensions that vary in relative significance throughout the process of the self-organizing dialogue (see Table 1). These five tensions were originally developed by Hammond, Anderson, and Cissna for a forthcoming article, “The problematics of power in dialogue,” in Communication Yearbook, Vol. 27. They have been expanded for this article to include a link to issues within complexity theory. Note that the social issues of identity, outcome, meaning, voice, and field are linked to complexity issues that point toward self-organization.

Table 1

The social issues in dialogue

Social issue Tension Self-organization
Identity Self and other The system defining itself
Outcome Content and process The system describing itself
Meaning Coherence and incoherence The system marking disorder
Voice Monovocality between mutuality The system marking order
Field Convergence or The system redefining itself


The first permanent tension in dialogue concerns the relation between self and other and falls under the label of identity. In a social system that is self-organizing, the communicative function of self-definition is always at play. In communication literature the label is identity, both individual and corporate (Burke, 1968). Communication, philosophy, literary theory, and psychological theories often suggest that self-knowledge and identity construction are possible only through the process of social interaction.

Burke described identification not only as the process of distinguishing one’s self from others but as the coequal process of creating a relationship with others. He argued that the formation of identification is what separates humans from animals. Deleuze (1990) believed that all communication begins with “I,” and that the individual is always comparing their central material condition with others. Buber (1958: 28), however, claimed that “Through the Thou a man becomes I.” He says that the tension between the self and the other leads to reflexivity, self-reference, and self-implication. This is the tension that calls the individual to enter into a social relationship through dialogue. The negotiation of self-identity and group identity is both a motive for and a product of dialogue. Identity is continually negotiated between the individual and the whole. The tension between self and other is therefore essential for communication and self-organization to occur, lending credence to those who believe that diversity defined in any terms (intellectual, cultural, racial, gender based, etc.) is essential for healthy dialogue. It also suggests that questions that negotiate difference, such as racism and sexism, are permanent.

But how does this “work” in dialogue? Organizational development consultant Marvin Weisbord (1992) seeks participants in his designed dialogues who will ask probing identity questions. He says that stakeholders who are concerned with their relationship to the group lead group members to question their own assumptions. To innovate and lead the group in new directions, says Weisbord, group members must come to see themselves differently. Simply put, new identities create new forms of organizing and different outcomes.


Different cultures define the process and content boundaries for dialogue differently (Kincaid, 1987). This is particularly true when defining what one might suggest is the starting point for dialogue. Hammond (1997) found that structured dialogues in culturally complex groups contain language that negotiates both content and process, but that one usually dominates at different points during the dialogue. For example, in some Asian cultures process must be made explicit before issues of content can be raised, while in some European cultures the content issues must be made explicit before issues of process can be discussed. These differences suggest that while the dialogic tension between content and process is permanent, it is also essential for social self-organization to occur.

Hyde and Bineham (2000) distinguished between content-privileged dialogue and process-privileged dialogue. The first, Dialogue1, binds participants to specific practices and presumes that they will engage in active listening and participative decision making across their differences. The success of this kind of dialogue can be determined by how well the participants address content issues. Dialogue2 is less predictable, less driven by content, and is process oriented. It engages us in unpredictable processes of self-implication, communing, and flashes of self-other insight.

One way to think about this distinction is to consider what facilitators can do to help dialogue develop. In Dialogue1, facilitators teach skills such as active listening, for example increasing the likelihood of careful consideration of positions among communicators. In Dialogue2, facilitators might primarily think of their task as clearing a space within which moments of dialogic encounter or insight might emerge with as few obstacles as possible. In either case, the tension between content and process leads to communicative behaviors that unite and divide the group around process orthodoxy and content outcomes. Groups socially selforganize as the dialogue leads them to consider what they want to do and how they want to do it.


In every social environment there is tension between the known and the unknown. Like the proverbial caution sign on the highway, “Watch for falling rocks,” every social system marks common coherence and incoherence. Palmer (1998) suggests that all things are known in the context of a community. It is this permanent tension between coherence and incoherence that demands a dialogue that leads to social self-organization.

Weick (1979), Berger and Bradac (1982), Berger and Calabrese, (1975), Shannon and Weaver (1949), and many others have described communication as a process of uncertainty reduction. Weick (1995) claims that we organize in order to create “sensemaking” environments that have a workable level of certainty. Organizations create a coherent environment where we act with some confidence and predict the effect of our actions. But Weick acknowledges that this is an ongoing process that, while desired, is never fully accomplished. Although each of these theorists rightfully argues that uncertainty reduction is a motive for individual communication, our position is that in dialogue the tension between coherence and incoherence is shared. Consider Tannen’s (1989: 152) description of pleasing conversations that lead to a sense of coherence: “The experience of a perfectly tuned conversation is like an artistic experience. The satisfaction of shared rhythm, shared appreciation of nuance, [and] mutual understanding … surpasses the meaning of words exchanged … It gives a sense of coherence to the world.”

Of course coherence, as we achieve it, is temporary and not always shared simultaneously with other participants in a dialogue. It forms the motivation to communicate in dialogue, and it is a hard-won moment more often than an enduring state of clarity. What is enduring, even permanent, is the tension between coherence and incoherence; that is, the rewarding and clarifying ends of dialogue, which can provide the impulse toward dialogue in the first place.

Hayles (1999) sees a similar tension in the semiotics of virtuality when she describes tensions between presence and absence, disruption and replication, and pattern and randomness in a self-organizing meaning system. For the participant in a dialogue, each of these tensional qualities is manifest as relative coherence or incoherence. Absence—physical or mental disruption—and randomness lead to incoherence. Presence, replication, and pattern lead to coherence.


The social system does not merely place a caution on disorder, but also marks order by certifying that some ideas have privilege over others. The privilege may come as a result of hyperdemocratization (a majority believe this, so it must be true), scientism (scientific experts certify this, therefore it is true), or some form of orthodoxy. But a majority population implies a minority view, just as experts who separate themselves from the average citizen imply isolation. Orthodoxy, whether religious or otherwise, invites disbelief. Each of these positions has an inherited, permanent tension that exists between the forces of mutuality and hegemonic monovocality.

Bohm (1990) tells us that dialogue leverages this tension by moving the group toward shared meaning. Others (Isaacs 1999; Weisbord, 1992) believe that dialogue moves us toward collective action. However, once the collective has formed it becomes controlling and resistance then becomes part of the self-organization process. Bakhtin (1981, 1986) said that a monologism, or single voice privileged over multiple voices, is needed to clarify the human condition. Holquist (1990: 24) told us, “in Bakhtin, there is no one meaning being striven for: the world is a vast congeries of contesting meanings, a heteroglossia so varied that no single term capable of unifying its diversifying energies is possible.”

This kind of monovocality stands in tension with mutuality described by Cissna and Anderson (1998) in the Buber-Rogers dialogue. They say that the kind of mutuality articulated by Buber, Rogers, and other theorists of mutuality leads to tension, dialogue, and then unexpected meaning. It emphasizes an individual’s awareness of the uniqueness of others and encourages authenticity. But it does not require the renunciation of roles nor the full disclosure of all personal thoughts.

An example of this is the Quaker clearness committee. This is a dialogic form of counseling when a person who is trying to make an important life decision calls together several trusted friends and family. They will sit in a circle for several hours as those who have been asked to counsel ask only questions of the person who is making the life decision. The questions may not contain prescriptions, such as “Why don’t you try this?” (Palmer, 1998).

This process lives in the tension between the monovocal and the mutual. It starts with a monovocal description of the question coming from the person who wants help. It guards the monovocal position by allowing the counselors only to ask questions. However, it evokes mutuality by including the clarity of the questioners and by stirring the wisdom that may lie deep within the person who is trying to make the decision. Evidently, a clearness committee is an attempt to leverage the forces of dialogue and lead a person to find order within their lives, even if the order is temporary.


Coherence is temporary, but it is critical because it allows the social system to converge on collective rather than random actions. Over time, social systems that are defined by order will need to redefine themselves. Social selforganization through dialogue is a nonlinear, communicative practice that involves emergent mutuality. As the collective intent of certain groups demands convergence on intended outcomes, a dialogic field of meaning acts much like a “strange attractor,” giving the group’s intention a sense of momentum. The vision, collective abilities, and experience of the group create boundary conditions that limit the outcomes, but human agency also opens up the system to infinite possible outcomes within those boundaries. The key question is whether the field is converging on a problem or series of problems, or whether it is open, waiting for a mutual direction to emerge.

The convergent-emergent tension is found in the literature of deliberative democracy theory. Habermas (1992) says that people come to a dialogue ready to hear others’ arguments and subject their own arguments to the same criteria of reasonableness and clarity as the positions they originally dislike. This creates communication that builds common ground and increases the chances of consensus. Consensus, reached through deliberation and verbal competition, appears to be the normative goal of convergent dialogue. Such convergent assumptions virtually mandate that unity, in the sense of shared meaning, is “the” goal of dialogue, not merely “a” goal of dialogue. They also keep other important voices out of the dialogue. The task in dialogue is not to submerge voices if they don’t reflect the “right” kinds of rationality, but to do what we can to bring them to the surface as contributors. The convergent-emergent tension is a reflection of the proceduralist-pluralist tension confronting a wide range of public and governmental decision-making situations.


Early in this article we described the characteristics of communicative self-organizing systems by describing a social system’s need to define itself, describe itself, mark order, mark disorder, and converge or emerge on action. We describe these as tensions because a social system is never without some dialogue-creating dispute related to these tensions. To be sure, the dialogue can occur on a variety of levels, including small groups, communities, societies, and even with oneself. We describe these dialogues as social self-organization because they give us, as subjective human beings, our sense of identity, the expectation of certain outcomes, the assumed meaning of communicative acts, a pretense of voice, and a sense of social situation or field.

The dialogic self-organization around identity moves toward establishing the self and the other, making clearer the roles of the participants. Dialogue around identity may lead to a re-entrenchment of identity roles or to newly established roles. The tension in dialogue between content and processes is also permanent and leads to self-organization. The tension between coherence and incoherence is required to begin dialogue (Kogler, 1996). However, as dialogue continues and fills in the limits of its own meaning, emergence can take over and move the group toward incoherence. Isaacs (1999) describes this as instability. Wheatley (1992) describes it as a productive, localized “chaos” that creates an opportunity for participants to let go of previous assumptions. We suggest that declarations of coherence are essential for self-organizing systems because they allow the system collectively to move forward with confidence. As US President George W. Bush declared coherence on an aggressive foreign policy that will force the disarmament of Iraq, he caused others to demand additional consideration. Congress, the United Nations, and the governments of other nations deliberated carefully to see if they would or could align with his position. The declaration of coherence created an opportunity for dialogue even as it was described as an imposition.

Kogler (1996: 84) argues that “dialogic understanding can no longer proceed from the idea of a universal consensus … it must … make present one’s own constraints through an understanding of the other, and of gaining knowledge of certain limits of the other through one’s own perspective.” We suggest that coherence and incoherence are ever present in the dialogic experience, surfacing and resurfacing at different times.

Finally, we consider voice in the self-organizing mix. In some dialogic situations, in a given time period, the monovocal voice, according to Hawes (1999), sweeps away any opposition. He argues that a single, privileged voice can be tyrannical and oppressive. Ellsworth (1989) says that some populations that are raced, classed, or gendered are so inherently unequal in their identity and access to the process that real dialogue is impossible. Nevertheless we argue that eventually, perhaps over generations, the system will self-organize and self-correct through the forces of mutuality. This was the hope of Freire (1970: 76-7), who said:

Dialogue is the encounter between men, mediated by the world, in order to name the world. Hence dialogue cannot occur between those who want to name the world and those who do not wish this naming—between those who deny other men the right to speak their word and those whose right to speak has been denied them.

In recent years there has been a great deal of discussion about selfsustaining organizations. It is clear that organizations that can self-correct toward more ethical behaviors need to sustain the forces of selforganization through dialogue. While there is a danger in the tyrants described by Freire who will keep us from dialogue in order to control the naming of the world, there is a more common danger in those who see dialogue as inefficient. In doing so they deny us access to our identity, pushing us toward a specified coherence spoken by a privileged voice. But more importantly, they deny the system the opportunity to self-organize and hopefully to adapt to our social environments in a more friendly and ethical fashion.