University of Wisconsin Center for Cooperatives
THE MICROPROCESS FEATURES OF
For many years, the literature on power and power relations paid little attention to what we might call the micropractices that constitute the very exercise of power—that is, control. In recent years, however, both within the discipline of communication studies and in cognate fields of the social sciences and the humanities, scholars have come to ask, How is it that power is exercised in a given situation or a given case (Deetz, 1995; Mumby & Stohl, 1991; Tompkins & Cheney, 1985)? In other words, how does democracy manifest itself in everyday organizational interaction? Communication studies can offer a great deal of insight into the microprocesses or concrete behaviors that constitute "workplace democracy" or "employee participation" (Cheney, 1995; Glaser, 1994; Harrison, 1994; Stohl, 1995a). In fact, participation may be usefully described as a special case of communication that meets a number of specified criteria: say, for example, relatively equal involvement, "deep" engagement in the issues at hand, and the possibility for revision or modification of the very system of participation (Stohl & Cheney, 1997). With these concerns in mind, it is crucial that we look carefully at the range of organizationally sponsored systems of participation or interventions, sorting through their substantive differences in terms of communication and democratic practice.
A recent study of the Grameen People's Bank of Bangladesh explored in depth what democracy and autonomy mean for female members of this large and successful grassroots organization (Papa, Auwal, & Singhal, 1995). The researchers note that although the specific interactional practices of the organization are highly democratic and empowering in certain respects, they also serve to constrain and burden members with a powerful and demanding vision and extremely high expectations for performance. This is just one case in which the enhanced democratic "control" brought about by new organizational forms can be double-edged in actual practice.
An examination of the micropractices of avowedly democratic organizations and those associated with democratic interventions within organizations can help to illuminate this issue and enable us to understand more fully what different models of workplace democracy mean in practice. At the same time, however, we must consider in detailed terms what it means for an employee to be engaged deeply in participative processes. We are treating employee participation as a special case of communication. However, we still must ask at what point an employee's contributions to discussions, to decision making, and to the ongoing and significant activities of the organization constitute meaningful participation. And we must recognize that even in an organizational environment that is genuinely democratic, there will be important differences in the persuasive abilities and skills of members (Murder & Wilke, 1970). Finally, a variety of other factors can affect group-level or organization-wide participation; these include compatibility of work schedules, information resources, member motivation, and leadership (Hirokawa & Keyton, 1995). Clearly, we are seeing employee participation and workplace democracy as beyond the normal, expected level of coordination that typifies organizational experience. In the remainder of this section, we characterize some of the issues of communication roles, discussion-related behaviors, and the bounds of discourse.
Making an organization more democratic involves much more than an alteration of its structure. An essential element of organizationally sponsored programs of participation is a change in organizational member roles and corresponding changes in patterns of communication. Although not explicitly discussed in extant research, role changes seem to function as a basis for a wide variety of other changes in microprocesses that are associated with the move from a traditional to a participatory organization (Schonberger, 1994). To varying degrees, programs to enhance participation ask employees who have traditionally functioned in follower roles to adopt participatory roles and ask managers to become, to one degree or another, partners or coaches or facilitators. The results of such a transformation can be the development of more dynamic communication networks and a shared sense of the "big picture" by organizational members (Stohl, 1986).
The literature on leadership is especially pertinent in addressing the question of how the micropractices of democracy affect employees and managers. The well-known writings on "transformational leadership" (Burns, 1978) emphasize a form of leadership that inspires others, engages them, stimulates them intellectually, and responds directly to their needs and wants. Many authors have applied Burns's transformational leadership style to the work-place (Hater & Bass, 1988; Kotter,1995;Tichy& Ulrich, 1984), and the style has been found to correlate positively with perceptions of leader effectiveness, employee productivity, and employee satisfaction with work.
Coaching is a leadership style that was recently rediscovered by organizations as a means of promoting workplace democracy. Although the term coaching originally described a manager-subordinate relationship similar to that of master-apprentice, it later came to mean coordinating the efforts of the whole team and determining what each member must do in order to ensure the best performance of the team. Today, coaching is considered to be a leadership strategy used to encourage maximum performance from each subordinate (Evered & Selman, 1989). Coaches strive to encourage group-based problem solving, the exploration of issues and diverse views, and an open and trusting communication environment. Coaching is a type of interaction between managers and employees that may be described as "a people based art that focuses on creating and maintaining a climate, environment, and context which enable/empower a group of people to generate desired results, achievements, and accomplishments" (Evered & Selman, 1989, p. 17; emphasis deleted). According to this model, managers create this climate, environment, and context through their interactions with others.
Building on the coaching metaphor, research on leadership has also explored the context of team-based organizational structures (such as semiautonomous work groups, outlined above). Manz and Sims (1980, 1984,1987) have examined the specific behaviors required and performed within the paradoxical role of the "unleader," the group facilitator who remains a coequal with others. In such team-oriented situations, leaders exchange their roles of motivator, trainer, and decision-maker for those of liaison, "connector," and mediator. Barry (1991) suggests a distributed leadership model that views leadership as a series of roles that can be adopted by any group member. Many of the activities that Barry identifies (e.g., getting acquainted, surfacing differences, presenting information to outsiders, summarizing positions, developing goals and vision) require particular communication skills, and each activity serves to enhance group work by facilitating group members' ability to work together and to accomplish tasks.
As the two examples above demonstrate, much of the literature on teams focuses on the alteration of management behavior, rather than on the changes required of employees. One important exception is the work of Barker and his colleagues (Barker, 1993; Barker & Cheney, 1994; Barker & Tompkins, 1994) that examines the implementation of self-directed work teams in a high-tech manufacturing company. In particular, these researchers found that the transformation of the follower role did not take place in a linear progression, but rather followed a two-steps-forward, one-step-back pattern of adaptation. For example, when faced with a crisis, the team members first asked someone what they should do, a habit learned in the formerly autocratic organizational structure.
Researchers who have addressed leaders" communication styles stress that leaders must demonstrate their commitment to the organizational vision and goals through both words and actions (Clement, 1994; Richmond, Wagner, & McCroskey, 1983; Senge, 1990). Other writers specifically emphasize the importance of nonverbal communication to the success of implementing and maintaining organizational programs such as job enrichment and gain sharing ((3runig, 1993; Kouzes & Posner, 1987; Remiand, 1981, 1984). A leader may talk the party line, speaking in favor of a democratic, participative workplace, but may communicate dislike for such policies through symbols, gestures, and other nonverbal means. Fairhurst and Sarr (1996) suggest that leaders should model desired behaviors for their employees and employ "framing" (or the inspirational management of meaning) to help employees understand their role(s) in the democratic workplace.
Whether leadership is associated with a team or with a separate manager does not seem to be as important as the attitudes, communication behaviors, and roles adopted by members of organizations trying to encourage participation. Organizationally sponsored programs of employee participation are paradoxically structured by managerial objectives; that is, to what extent are employees required rather than allowed to engage in participative practices? Many researchers have worked in earnest to develop a model of organization in which participation and democracy are principal characteristics of the organization, and some observers offer feminist organizations as cases in point.
Feminist theorizing about organizations offers much insight into the "reinvention" of democracy in organizations, and certain feminist organizations provide excellent examples of leadership within the context of democratic organizations. Such organizations tend to be based on high member participation, have relatively flat hierarchical structures, and frequently emphasize practices of cooperation (Rodriguez, 1988). Feminist organizations generally strive to be democratic or even consensus based, exemplifying cooperative rather than competitive principles, and are supportive of their members (lannello, 1992; Pardo, 199S). In this sense, these organizations embody shared leadership.
However. explicitly feminist, egalitarian organizations may encounter special difficulties. Manshridge (1973), notes that participatory groups, including many avowedly feminist organizations, typically face three challenges: First, decisions tend to take more time; second, decisions demand more emotional involvement and vulnerability than they do in mainstream organizations; and third, participation is not equally available to all participants because they come to the organization with different knowledge, skills, and backgrounds. Claser (1994) has recast these three dimensions in explicitly communicative terms, arguing that emotional expressions can serve multiple functions in democratic egalitarian groups or organizations. Again, we are reminded of the need to examine the specific behaviors and patterns in the avowedly democratic organization, being aware of both advantages and disadvantages of particular actions.
Along with the strong connection between feminist theory and workplace democracy, there is a direct connection between feminist ideology and democratic micropractices. Generally, feminists treat the personal side of relationships as being extremely valuable, as the aphorism "The personal is political" expresses. Workplace relationships are seen in this same light. Feminist values, applied to the workplace, tend to emphasize working relationships characterized by mutual support, empowering behaviors, caring, cooperation, and fairness of treatment (Martin, 1990).
Feminist researchers also criticize the separation of the "public" work world from the "private" home (Maguire & Mohtar, 1994; Mumby, 1993). Traditionally, an employee is expected to "check her (or his) private life at the door"—that is, to leave non-work-related issues behind. Maguire and Mobtar (1994) argue that by recognizing both the private and public lives of workers, an employer can begin to value employees' entire reality, to promote a type of participation that acknowledges individual differences and yet does not completely "absorb" the individual into the organization. Levering and Moskowitz (1993) report on an example of this type of working relationship at Patagonia, a U.S.-based outdoor products company. An employee of Patagonia described the company as unique and remarked that mothers employed there are encouraged to breast-feed their babies at work: "A woman walked into a meeting and breast-fed her baby. She gave this great presentation. Everyone agreed with her point, and there was a baby sitting right there. It was so normal and healthy and natural that it just fit" (p. 342).
Although the study of leadership roles and types of behaviors reveals much about what both "democracy" and "participation" can and could mean at the level of organizational activity, a fuller examination of the microprocesses demands attention to patterns of organizational discourse. Here we refer especially to the ways certain linguistic devices implicate the "management of meaning" and therefore serve to delimit or expand the range of democratic possibilities.
Mumby (1988) has written extensively on this topic, usually under the rubric of power and control in organizations. He has demonstrated how narratives or stories in organizations can function to solidify individuals' roles, maintain status differentials, and limit forms of expression. Whitten (1993) similarly reveals how office tales can "keep employees in their place ." Whitten relates an often-told story in one corporation of how a dedicated employee was fired for even suggesting that his superior shouldn't worry about being a few minutes late for a meeting while the superior was driving over sidewalks and yards to get to an appointment with a CEO of another company. Such seemingly "interesting" stories at work can serve to dampen dissent, discussion, and dialogue, thereby inhibiting any real movement toward a participative or perhaps democratic work organization. Clair’s (1993) research on sexual harassment shows how the less powerful can participate in their own subjugation by using "framing" devices such as "reification" (e.g., "That's the way it is in our society") and "trivialization" (e.g., "It's not that big a deal").
With respect to language, discourse, and the microprocesses of participation, the ambiguities and vagaries of language itself deserve careful attention. Eisenberg (1984) calls attention to how ambiguity can be exploited as a strategic resource for communication and interaction in organizations. Following in this same line of analysis, Markham (1996) considers how ambiguous messages within the context of one organization's program to stimulate employee creativity and participation are strategically employed as a vertical control measure.
Finally, we return to a brief consideration of the nature, dynamics, and "evolution" of key terms such as democracy and participation. Not only do such terms, which are important in colloquial organizational discourse and not just scholarly discussion, elicit multiple meanings, but their dominant meanings in particular contexts can shift significantly over time. For example, Cheney's (1997a,1997b) research on the Mondragon cooperatives shows how a prevailing understanding of participation in terms of a right to contribute to decision making and even policy shaping is perhaps giving way to a managerially promoted demand for participation "at the level of one's job"— in practice meaning "the giving of one's all" to the job, team, organization, and ultimately the customer (as a point of reference). These examples of discourse in practice suggest the critical importance of delving deeply into the meanings of "democracy" and "participation" as well as considering more broadly what practices count as participation or democratic governance.
THE ISSUE OF VOICE IN
This criterion for assessment of the extant literatures asks, To what degree does a body of research treat the ways in which multiple voices are expressed, suppressed, repressed, co-opted, or ignored in the discussions that presumably constitute workplace democracy? The literature on power is especially well developed in responding to this question, in that it has ready concepts of interests, voice, hegemony, and so forth (Clegg, 1988; Lukes, 1974). However, as Scott (1990) has shown, these issues are by no means resolved definitively, even within the conversation of those who talk explicitly of power relations (in and about organizations): What appears to be "false consciousness" of majority and/or minority groups in some cases may simply he an adept performance of submitting publicly to a form of domination while maintaining an elaborate system of egalitarian relations among the comparatively disempowered as they act behind the scenes. Thus Scott questions the tradition of presuming hegemony of dominant ideologies by observing that subordinated groups—from slaves to secretaries—frequently have their own counter-discourses, their own hidden and emancipatory "texts."
The questions raised by the literature on power must be considered as well with respect to other literatures that would presume to say something important about the prospect for democracy at work. For example, how does each literature deal with questions of outright coercion versus subtle dimensions of shaping a discussion through the sheer definition of terms? The research on alternative organizations, we should observe, has not generally been very reflective about how its very ideals—for example, equality or solidarity or participation—can function as tools of domination by those who hold the keys to defining those values.
In this section we will consider the important matter of voice, reflecting not only on the ways in which it has been treated by a variety of research traditions (Hirschman, 1970) but also on the utility and limitations of the metaphor itself (Mumby & Stohl, 1996; Putnam, Phillips, & Chapman, 1996). Along the way, we will address the ways in which the idea of employee voice, in terms of open expression of ideas (Gorden, 1988; Haskins, 1996), can be related to the notion of "interests." And we will consider the communication-related constraints on the expression of voice by both individuals and groups.
Democracy extends simple participation in the workplace by ensuring that the individual has a voice, may express an opinion, that means something and has the potential for "making a difference" in the larger organizational context. In this sense, the Western rationalist-democratic tradition is strongly biased against silence, in contrast to, for example, many Native American cultural traditions.
Considerations of voice and power in organizations may be closely linked to the function of power structures in the workplace. As indicated above, structures enable and constrain certain democratic practices. On the overt or explicit level, democratization can be operationally defined in terms of the extent of genuine opportunities for dissent or discussion. For example, Gorden (1988) specifies different stances and corresponding communication behaviors associated with voice; these stances—ranging from "active constructive" to "passive destructive"—apply to the larger community as well as to work life within the organization. Among other things, Gorden and his colleagues (e.g., Kassing, 1996) have pointed out differences in employees' perceptions of their free-speech rights at work, both across different organizations in the United States and between different countries (Ewing, 1977).
The most dramatic form of dissent, of course, is whistle-blowing, the act of going outside the organization to express objections to organizational policies or practices (Stewart, 1981). In a sense, however, whistle-blowing can be taken as an indication that the system does not allow for dissent in everyday organizational practice. Thus it is important to consider multiple types, avenues for, and expressions of organizational dissent (Kassing, 1996), democratic tendencies of any organization.
On the level of ideology and socialization, the ways systems of discourse are shaped are central to our understanding of the range of expression. Mumby and Stohl (1991) analyzed an interaction of team members for their absence and presence in team meetings and gatherings. The work team took issue with a member who had recently been absent, and when the employee tried to explain his absences by referring to "stuff" that had come up, another team member responded, "Come on, there isn't other stuff" (p. 322). As Mumby and Stohl observe, the team forced private issues out of the workplace. The team allowed only those issues directly related to the team's work and tasks to be considered. Other "stuff," such as health or family concerns, had no place in this team and in this particular workplace. A democratic workplace, then, must consider and value various voices and the concerns that diverse voices may raise. The structures of such a workplace must be able to account for the needs of employees. In this way, the metaphor of voice is a useful critical tool.
Deetz (1995) has recently proposed a model that attempts to account for the largely suppressed voices of organizational members and constituencies. The model attempts to position the corporation as a forum for the productive airing of various constituencies' voices. Specifically, Deetz advocates a dialogic model of communication in which various groups of corporate stakeholders not only express their perspectives but also have equal opportunities to influence the decisions made in and by the corporation. Many of the changes that Deetz presents relate to efforts to democratize the workplace and the roles of communication in the democratic workplace: Each member should adopt the perspective of an owner, information should be readily accessible, structures should be shaped by those at the bottom of the organization, and interactive discussions and negotiations of values and ends should occur on a frequent basis. From this perspective, the very discourse that characterizes much of the organization is expansive and self-reflexive rather than closed and self-satisfied.
Standpoint feminism (Buzzanell, 1994) is relevant to issues of voice within the workplace. Standpoint feminism argues that women, like men, have a variety of experiences and thus have different "standpoints" or perceptions/opinions from which to view the world. Beginning originally with a critique of social institutions from the perspective of middle-class white women of the industrialized world, standpoint feminism has come lately to consider a variety of excluded or marginalized groups and what can be learned about the reform of society from their multiple viewpoints (Bulks & Bach, 1996). Standpoint feminism, in this sense, highlights the special understandings of the social order held by comparatively disempowered individuals and groups. From this ideological stance, a truly participatory, democratic organization should value, or at least recognize, the voice of each employee, whose life experiences may include the roles of manager, mother, wife, volunteer, victim of sexual assault, and so on. All of these roles or life experiences shape the views of this single employee and all contribute to her overall perspective on life and work (Gottfried & Weiss, 1994). As another example, a single man who lives alone may have a different standpoint on a day-care issue from that of a man who is helping to support five children. The organization should recognize both viewpoints in order to promote continued participation by these employees and to foster a supportive working environment. The essential point here is that a particular workplace issue can look different from a variety of points of view—especially when employees' viewpoints are based in lived experience.
The recognition of multiple voices can, however, pose problems within an organization. Feminist ideals of collectivity and respect are often lost in favor of getting things done. As Ferree and Martin (1995) observed in their study of feminist organizations, "Overtly feminizing efforts started with emphases on collectivity and consensus of various degrees, but over time they moved to more hierarchy or to representative rather than participatory democracy" (p. 138). Making sure that multiple voices are heard can take time, and, unfortunately, corporations typically equate time with money. This pervasive dilemma is not one that will be easily solved, but it is worth investigating further. Feminist theorist Joan Acker has stated: "I believe it will take radical transformations of the entire society, which we cannot yet imagine, to create conditions that will support alternative and humane forms of organizing. In the meantime, the feminist image of the nonoppressive organization can serve as an ideal against which to judge our actions" (quoted in Ferree & Martin, 1995, p. 141).
A more traditional view concerning voice has been aired by Conte (1986) in his examination of the effects of ESOPs on organizations' performance, productivity, and employee attitudes. This examination was conducted in response to the "participationist" argument that the positive effects of ESOPs on employees' productivity are due not solely to the profits gained by workers, but also to the employees' accompanying right to participate in decision making (or to voice their opinions). Conte asserts that although profits are certainly an incentive for workers to get involved in ESOPs, such systems would be ineffective if employees did not feel they had some say over the decisions their companies make.
Along these same lines, Kornbluh (1984) found that problems may arise when firms try to implement quality circles and QWL programs without also implementing a democratic form of management. Workers will be disappointed if they are introduced to such programs and their suggestions are not implemented by management. Kornbluh sees the current U.S. workforce as more resilient to authoritarianism than the previous generation of workers. Therefore, he argues (hat it would be ridiculous to spend energy implementing quality circles without also instituting a flatter hierarchy and allowing more worker participation and decision making throughout the organization.
Within the context of organized labor, voice as a practical construct refers to autonomous practices of speech that are braced by institutional guarantees. The right of free speech guaranteed to U.S. citizens by the First Amendment largely disappears within the context of the workplace; the Bill of Rights does not protect workers from private sector abuses. Laws that govern speech-related rights within the context of organized labor are thus rooted in an amazingly complicated mixture of Supreme Court rulings, National Labor Relations Board rulings, union constitutions, and agreements made during the course of collective bargaining.
Though many constraints exist within factories and offices that stipulate the kinds of talk appropriate or inappropriate in the workplace, mechanisms do exist that allow workers to channel discontents stemming from conditions of work. Historically, the routine guarantor of voice in the workplace has been the grievance procedure. Brody (1992) has characterized the grievance procedure as the most important mechanism for indicating blue-collar workers' resistance to poor working conditions. Until the wide popularization of company-sponsored employee participation programs, the grievance procedure was usually addressed in labor-related scholarly literature as the one structural guarantee of voice (Chamberlain, 1951; Cooke, 1990; Elkouri & Ellcouri, 1980; Herman, Kuhn, & Seeber, 1987; Rothenburg & Silverman, 1973).
Since the mid-1980s, labor scholars rooted in traditions of industrial relations research have discussed voice within the context of the increased participation afforded by the seemingly democratic structures and practices of quality circles and work teams. These scholars note that increased participation often lends to workers a sense of increased workplace control (Marks, Hackett, Mirvis, & Crady, 1986) and crafts for them a perceptual structure of equality of power between labor and management personnel (Stuart, 1993). However, these findings are countered by the publication of occasional, but strong. dissenting research. Kelley and Harrison (1991) found that participation programs do not lead to higher productivity and, importantly, that the democratizing benefits of such programs disappear unless they take place in a union setting.
Ultimately, each presumed case of workplace democratization needs careful scrutiny with respect to such dimensions as (a) the range of issues about which participants may speak, (b) the extent of actual influence by employees through their exercise of voice,.and (c) the levels of the hierarchy at which meaningful voice is possible (Bernstein, 1976; Cheney, 1995; Miller & Monge, 1986; Monge & Miller, 1988; Strauss, 1982). Still, any "textual" emphasis on what is said in democratic discussion and participation must be complemented by an exploration of the "counterfactual"; that is, what sorts of expression might be possible within an alternative or different system? In considering both empirical realities and model alternatives, communication research can speak specifically to the issue of voice in terms of what speech practices "count" as meaningful democratic expression and how they can best be promoted and protected.
ADVERSARIAL AND CONSENSUS-BASED IMAGES
The issue of adversarial versus consensus-based models is undoubtedly one of the most important in democratic theory (Mansbridge, 1983). It implicates several concerns, such as organizational size, group homogeneity, and value homophily, and the very definition of democracy itself. And, even more broadly speaking, we encounter the question of possibilities for real consensus, rational dialogue, and a democratic order. Within communication studies, perhaps rhetorical theory has been most directly concerned with this issue. Burke (1969) has offered a vision of rhetoric that is explicitly cooperative rather than oppositional. In terms of the literatures we are reviewing for this project, studies of contemporary organized labor, as well as analyses of feminist and other alternative organizations, are especially relevant. Organized labor has traditionally been a promoter of workplace democratization (Freeman & Medoff, 1984). However, in reaction to some current specifications and enactments of that term, segments of organized labor have argued against a few organizationally sponsored systems of participation (Parker & Slaughter, 1988, 1994). Of course, such opposition is understandable in terms of the union-busting motivation of some programs (Crenier, 1988). Braverman (1974) questions seriously whether such programs can be anything but co-optive within the status quo of labor and industrial relations.
Our chief concern here is how democratic systems ought to be constructed (Whalen, 1997). Some would say that only an adversarial model such as Minsky's (1971) can be truly democratic, pointing to the need for an institutionalized force of opposition or resistance within any organizational system. Other observers, stressing what Mansbridge (1983) would call a "unitary" model of democracy, would argue that a genuinely democratic system ought to strive toward consensus (see Sheeran's, 1983, study of the Friends Church). Obviously, these issues also implicate the literature on power in organizations, especially considering the extent to which true consensus is seen as achievable. How the tension between adversarial and consensus-based models of democracy is resolved in communication practice is just as important as how it is conceived within the various bodies of research covered by this essay. As Mansbridge explains, the crux of the question is in how best to represent the interests of a group or groups.
Therefore' in this section we will consider what each of the literatures reviewed has to say about balancing concerns for oppositional arrangements in discussions versus seeking consensus. For example, some current theories of puhlic relations tend to emphasize "symmetry" in relations between and among organizations (Grunig, 1992). Although such an impulse is democratic in seeking to engage as many organized groups as possible in discussions about salient issues of the day, in practice the model tends to be equated with sheer agreement. In the literature on alternative organizations, we find similar treatments of consensus within the organization, essentially begging the question of how disagreement is conducted within the context of an organization that is expressly committed to unity of commitment and homogeneity of values (Westenholz, 1993). Three key questions then emerge: When is unitary democracy possible? How can an organization that embodies a unitary vision of democracy prescribe and truly grant a role for oppositional groups? How can an organization built upon an adversarial model of democracy arrive at a sufficient degree of consensus in order to thrive and maintain the coherence of its governing system? The answers to these questions are interrelated and entail important communication-related concerns.
In her anthropolitical investigations of an array of community organizations, Mansbridge (1983) identifies size as a crucial factor in, or even determinant of, the possibilities for a thriving consensus-based democratic order. In this way, her investigations echo the pessimism of both Weber (1978) and Michels (1962), who both saw the growth of an organization as a significant impediment to democratic practices and the maintenance of an egalitarian ethos. However, there is an important distinction to be made here between, for example, a law-like or deterministic perspective (such as that of Michels and perhaps that of Weber) and the recognition of strong tendencies. Do we push for consensus whenever possible, assuming an achievable unity of interests and goals? Or do we structure an organization so as to ensure and protect the expression of different and even opposing voices? In a study of collectively oriented law firms, Hansmann (1990) takes the side of determinism, arguing not only that unitary democracy is achievable solely in groups of small size but also that it requires groups of highly educated and similarly trained professionals.
Certainly, the recent literature on power in and around organizations cautions us against presuming that real consensus is possible, especially over time. in even small organizations. Lukes's (1974) theory, in the tradition of recognizing social "hegemony" (particularly in terms of how potential minority voices are socialized out of the discussion), and Foucault's (1984) insistence on the importance of there being constant opportunities for the expression of minority voices "on the margins" are two prominent examples. As an empirical case, Westenholz's (1991) longitudinal studies of worker and other cooperatives in Denmark reveal ways in which small as well as large organizations can forcibly try to improve and maintain value consensus, thereby engendering paradox.
The literature on feminist-oriented organizations reveals perhaps the most confident school of research in terms of the prospects for democratic, consensus-based organizations. Feminist organizations are often characterized by design, as cooperative and consensus based. Cooperation is defined here as a willingness to help others rather than to act competitively against them. Cooperation is evident both within feminist organizations and in relations between feminist organizations and other organizations.
Rodriguez (1988) describes a battered women's shelter as being cooperative and nonhierarchical. In this organization decision making is by consensus, hiring is based on a commitment to social service and firsthand experience of the problem rather than professional experience, and the workers' salaries are all the same. Iannello (1992) reflects on the limits of consensus through studies of various women's groups with feminist commitments. Importantly, lannello recognizes that not all decisions in nonhierarchical or minimally hierarchical feminist organizations need to go through the full consensus process for the organization to maintain its value commitments. Specifically, only those decisions that are deemed vitally important to the group need to be based on consensus—a form of "modified consensus." Thus groups or organizations can decide upon the domains of decision making where consensus is most valued and those where it is thought to be indeed constitutive of the organization. This is consistent with Kanter's (1982) analysis of the appropriateness of thoroughly democratic decision making for various organizational situations.
Labor research and theory have also addressed this question extensively. As many labor organizers put it, Can any nonoppositional, nonpluralistic system of governance avoid the traps of co-optation and the silencing of important minority voices? This is why labor scholars Mandel (1975), Parker and Slaughter (1988, 1994), and others are so critical of organizationally sponsored systems of participation. Not only do such systems (e.g., quality circles or self-managed work teams) typically have limited scope; they also tend to grant more responsibility to employees without also endowing them with increased decision-making discretion. Whereas the Commission on the Future of Worker-Management Relations (1994) has pronounced the current period a new era in labor-management cooperation, Parker (1993) calls the typical enactments of this philosophy "management by stress."
In fact, the issue of whether or not unions should cooperate with management has been characterized as "the key strategic question for the American labor movement as it struggles to survive the 20th Century" (Banks & Metzgar, 1989, p. S). In the protracted and sometimes bloody history of labor-management relations in the United States, the central operating assumption has been the materialist assertion that labor and management exist
in a necessarily adversarial relationship. Further, organized labor has assumed that the different interests that drive labor-management negotiations necessitate a democratic system that operates independent of management influence.
A large body of contemporary literature characterizes the efforts of workplace democracy programs in terms of "ideologies of cooperationism." The central issue in this literature revolves around the question of who should control shop-floor knowledge (Bluestone & Bluestone, 1992; Parker & Slaughter. 19X9; White, 19X9). Organized labor holds that shop-floor knowledge should remain the sole property of labor, and that cooperation with management on related issues means the co-optation of labor's chief domain (Parker & Slaughter, 1989). Management often holds that shop-floor knowledge should be the province of both labor and management, and that many benefits accrue to workers in settings where such information is publicly discussed (Bahr, 1989; Swinney, 1989). This issue has important practical implications in teens of communication, in that information and control are closely linked in practice (Burawny, 1979).
There is little doubt that organized labor has entered a new era, one in which assuming a purely adversarial stance holds little payoff. As Bluestone (1989) notes: "While there is some nostalgic appeal to 'old style' adversarial unionism, the lesson of the cards seems compelling. The union, no matter how militant its stance, has little power to tame the global marketplace or for that matter reign in the multinational firm that moves its operations abroad or outsources its production to avoid the union" (p. 68). Put another way, "if it's the only game in town, then it may be necessary to play" (p. 69). The question is not whether adversarial or cooperative stances are preferable for labor, but rather what models of cooperation are more structurally democratizing and thus preferable to others.
In closing this section, we want to emphasize the essential differences between adversarial and consensus-based models of democratic practice while also recognizing their complementarily. Consensus-based images of organizational democracy often presume that authentic harmony is possible and achievable. Adversarial notions of democracy assume that multiple interests will necessarily be in competition with one another. Although it is likely that the size and nature of an organization, along with the preferences of its members, will point to one model as more suitable than the other, it is also true that each model can be seen as a corrective to the excesses of the other. Seen from this perspective, then, communication would take on multiple formal and informal roles within an organization that incorporate some elements of both consensus-based and adversarial models of democracy.
At the same time, however, we must remain mindful of the postmodern critique that calls into question the very meanings of consensus and the open expression of interests.4 We should not take adversarial or unitary models of democratic participation for granted; rather, we ought to "deconstruct" each
type to reveal the ways in which its operation in practice may undermine its own presumed goals. This is important because most of our models of democratic participation, whether consensus-based or adversarial in orientation, presume that training and engagement in rational discussion represent progressive movement toward democratic ideals (Mumby, 1997). However, as a number of theorists (who may be loosely clustered under the heading of "postmodernism") have shown us, there are real practical limits to our neo-liberal conceptions of democracy and to the ways democracy is manifest today: Consensus may be false, shallow, or stale (Phillips, 1996); fragmentation may be essential to democratic vitality (Foucault, 1984); more communication itself may be problematic (Baudrillard, 1983); our frequent and rapid-fire methods of measuring public opinion (e.g., of consumers) may actually discourage deep forms of participation (Laufer & Paradeise, 1990); and incipient or overt "corporatism" can privilege the voices of the best organized and most-resourced bodies, even in an apparently open discussion (Cheney & Christensen, in press).
We place concerns about control here for three specific reasons: (a) because the verb to control is more suggestive of action than is the noun power (Tompkins & Cheney, 1985), (b) because we observe both that there exists an identifiable literature on power and that issues of control have surfaced in a variety of areas of scholarship, and (c) because the question of control readily suggests the issues of horizontal versus vertical aspects of organization and reorganization (Mulgan, 1991). Self-directed work teams may promise an increase in autonomy for the employee, yet control may be strongly exercised by the group over the individual (Barker, 1993; Barker & Cheney, 1994). Another example is the debate over whether computer-mediated communication (and other related forms of communication technology) will lead to radically democratic structures or to new forms of centralized control (Mantovani, 1994; Sproull & Kiesler, 1991). Part of what is interesting about this debate is that we are observing emergent communication patterns and structures even as critics and analysts are arguing about where they are headed.
In this section we will examine how control is manifest in the ongoing interactions and activities of work. At least three important issues emerge here. First, we must consider the extent to which horizontally organized systems actually grant enhanced control to individual members of the organization. Second, related to Tannenbaum's (1983, 1986) "control graph" theory, we must consider the extent to which the total control in a system is finite or expandable. What do transformations in the structure of an organization toward greater democracy mean for the total amount of control, and how would we explore such a question in terms of the interactions that constitute both communication and the organization? And third, how do language and other symbol systems operate in practice so as to constrain or even undermine presumably democratic practices at work?
Edwards (1979) has argued that to the extent that employee participation programs do not reshape the structures of power in the workplace, these programs remain a tool of management. This idea fits well with much of the empirical evidence provided by a variety of organizational scholars (Barker, 1993; Manz & Angle, 1987). Although we believe that only deep member involvement may significantly reshape the workplace, we recognize that even democratic practices can become structural in nature, enabling certain practices while constraining others (Giddens, 19X4). Control becomes an important issue on both the "local" level (e.g., the implementation of self-directed work teams) and the broader organizational level. In both domains, what first appears to be enabling may actually be substantially constraining. Moreover, the various levels of organization can interact with one another in some surprising ways, such as when a team-oriented ideology comes to supersede organizational concerns (Sinclair, 1992). The question of control is evocative of surprise, contradiction, and paradox (O'Connor, 1995; Stohl & Cheney, 1997: Wal ker, 1996; Westenholz, 1991).
Perhaps the most relevant paradox is what Stohl and Cheney (1997) call the paradox of design. This refers to efforts to build the "architecture" for participation in a largely top-down manner. This raises the question of the degree to which workplace democracy can or should be a managerial prerogative. Still, it may be useful to consider under what circumstances a participatory work climate (Reading, 1972) can be fostered or facilitated, with a certain acceptance and perhaps creative transcendence of the paradox.
Russell's (1985,1996) studies of diverse forms of worker co-ops (including Israeli kibbutzim, San Francisco's taxi driver associations, and emergent employee participation in Russian private firms) reveal the importance of looking closely at the economic and social arrangements characterizing any avowedly democratic firm. In his studies of ESOPs, for example, Russell (1984) surprisingly concludes that under certain conditions, economic ownership (e.g., in the forte of guaranteed stock options) can actually result in greater political and social powerlessness for employees, because the symbolism of property ownership can divert workers' attention from labor-oriented interests (such as improvement of working conditions and enhanced participation in decision making).
Probably no other type of organizationally sponsored program of employee participation or workplace democracy has been promoted more in recent years than self-directed, semiautonomous, or self-managing work teams. Katzenbach and Smith (1993), among others, celebrate the transformational possibilities of emerging team-based structures: flatter hierarchies, greater coordination, heightened participation, and, of course, increased productivity for the organization. Consistent with this perspective, Snyder and Graves (1994) summarize the philosophy of empowering employees, both as individuals and as team members. Specifically, they argue that leaders cannot force employees to change; doing so produces only short-term change. Leaders must adapt their visions of the future to employee suggestions when appropriate. Empowerment involves much more than just delegating tasks to employees:
Empowerment means giving employees jobs to do and the freedom they need to be creative while doing them. It means allowing employees to try new ideas, even ones that have never been considered or that have been previously rejected. It means allowing them to experiment and fail on occasion without fear of punishment.... [Lenders! should establish an understanding with employees about the risks they are willing to take in the experimentation process. (p. 6)
It is important to observe, however, that even from this comparatively employee-centered perspective, empowerment is seen as an organizational imperative—defined, interpreted, and applied from the top.
At the same time, there are promoters of work teams who caution organizational leaders against surrendering too much control, urging them to communicate clearly to team members the parameters of the teams' autonomy and organizational jurisdiction. In this vein, Simons (1995) argues for the steady maintenance of leaders' roles both outside and within work teams. In fact, Simons details several strategies to help leaders maintain control: Diagnastic control systems allow managers and employees to observe progress toward predefined performance goals. Belief systems empower employees at all levels by communicating the company's primary goals and encouraging employees to look for new opportunities. Boundary systems are the limits that must be placed on employee efforts to try new ideas and take risks, and might include minimum standards of performance, companies or ventures to be avoided, or practices that are not allowed. Finally, interactive control systems help leaders to involve themselves regularly and personally in the decisions of their subordinates. More specific to a team's functioning in terms of group interaction is the potential for greater horizontal (Craham, 1993), "endogenous" (Mulgan, 1991), or peer-based control (Barker, 1993). Although team-based structures are designed and implemented largely with a "postbureaucratic" motivation, they may well take on strikingly bureaucratic features, as demonstrated by Barker's (1993) study of a midsize high-tech firm. Employee team members in this case came to police one another in terms of strict adherence to rules for attendance and performance, effectively substituting group control for certain dimensions of vertical control over work processes and employee behavior. Especially remarkable in this study was the extent to which team members actually internalized these rules, norms, and ideas about appropriate individual performance (see also Sewell & Wilkinson, 1992).
Another dramatic example of increased control over individuals' work behavior comes from Manz and Angle ( I 9X7), who examined a different work context that consisted of individual insurance representatives who had formerly experienced a high degree of autonomy, making their own schedules and reporting on their own activities. Manz and Angle found that in this case the implementation of work teams not only reduced autonomy for employees but also diminished job satisfaction and harmed customer service.
Perhaps the most vivid published analysis of control in a team-based organization is provided by Graham (1993), who offers an "insider's" look at work processes and social relations in a Japanese "transplant"—specifically, a Subaru-lsuzu plant in West Lafayette, Indiana. In general, Graham found that the Japanese managerial system exerted a rather thorough measure of control over worker behavior. In fact, he felt that only truly collective resistance on the part of workers could overcome such complete control. When an individual resisted by leaving his or her team's place on the assembly line momentarily, production was accelerated by other team members upon that person's return, as a form of peer punishment. In order to analyze the various aspects of control, Graham divided the production process into five social components: (a) the preemployment selection process; (b) orientation and training for new workers; (c) the team concept; (d) the philosophy of kaizen, or continuous improvement; and (e) attempts at shaping a shop-floor culture. The team concept was implemented on three levels, stressing self-discipline, peer pressure, and the intervention of the team leader. The entire system was set up so that the failure or error of an individual led to "punishment" of the team, resulting in even more intense peer pressure and oversight by the leader. Kaizen was implemented so that workers' ideas were appropriated by management, and management decided if, where, and when to employ the ideas. The shop-floor culture was shaped and reinforced through a stress on teams and on an overall team culture; in this way, the company hoped to foster both team-based and organizational identification. For example, everyone wore the same uniform and was referred to as an "associate." The team metaphor was promoted at all levels.
Graham's (1993) study, and the others discussed in the preceding paragraphs, raises the issue of total control in an organizational system. That is, we can examine the implementation of work teams (and other organizationally sponsored systems of participation) not only with respect to the group's exercise of control over the individual, but also with regard to the overall amount of control in the organizational system. In fact, Coleman (1974) argues that although the total amount of control for "natural persons" (unorganized or unaffiliated individuals) and for "corporate persons" (or organizations) has greatly expanded in modern times, the proportion of individuals' control has declined relative to organizational power.
On the organizational and institutional level, Tannenbaum (1983, 1986) has devoted significant attention to the question of what changes in organizational structure or practice can increase the total amount of control in a system. In his earlier work, Tannenbaum ( 1983) asserted that the use of employee participation programs would increase the total amount of control in an organization. Control should not be considered as governed by zero-sum logic, but rather as an expandable dimension of the organization. As employees increased their control of, and responsibility for, organizational processes, the control present in the organization grew. In a more recent article, Tannenbaum (1986) draws on existing literature to demonstrate his earlier argument. Studies of participative/nonparticipative organizations in the former Yugoslavia, Germany, Japan, the United States, and Canada tend to show that the amount of control in participative organizations is greater than the amount of control in nonparticipative organizations. However, Tannenbaum's (1986) idea of participation is left somewhat ambiguous and does not necessarily feature the deep level of involvement that we might posit for true workplace democracy. In some of the examples that Tannenbaum cites, participation programs increased the amount of control in the system while at the same time highlighting the differences in control between managers and subordinates.
With respect to organizational control, one of the distinctive contributions of communication-centered studies of workplace democracy and employee participation is the specification of the precise nature of participatory constraints, possibilities, and activities. On a general level, Mehan (1987) reminds researchers to attend closely to the empirical connections between micropractices that may produce power inequities (e.g., socialization and education). Willis's (1977) detailed study of the socialization of working class youths in Great Britain shows vividly how a variety of institutional experiences can together serve to reproduce the social order and ultimately limit the participation of individuals in professions, organizations, and the larger public sphere. Knights and Willmott (1987), in their studies of service industries, show how "organizational culture" can sometimes be used as a managerial strategy, a tool for control, that limits rather than widens possibilities for understanding and participation by organizational members. Similarly, Kunda (1992) explains how the normative control associated with a strong organizational culture in a high-tech firm can produce severe tensions between organizational pressures and individual aspirations. Tompkins and Cheney (1985) argue that contemporary reliance on "unobtrusive control" strategies in organizations—especially the fostering of employee identification and the internalization of corporate values—can be double-edged in that it may strengthen social bonds that both enable and constrain the individual. In an extensive study of the U.S. Forest Service, Bullis (1991) considers specific communication practices, such as information dissemination, that serve to inculcate organizationally preferred values and thereby make control over individual decisions more likely. And Alvesson (1996) examines a business informational meeting to show how subtle dimensions of power are operative even in apparently value-neutral activities. All of these studies remind us of the importance of analyzing socialization (in its various forms) as we consider the potential for and actuality of workplace democracy.
We close our review with a brief discussion of the views of one of the most prolific writers on the subject of communication, power, and democracy, Stanley Deetz. Deetz (1992, 1995) considers the organizational domain for democracy both in terms of ideas of dialogue and the public sphere (Habermas, 1989) and in light of constraining notions of communication and its capacity for marginalization (Foucault, 1984). Although highly skeptical about the potential for real democracy through organizationally sponsored systems of participation, Deetz sees hope in such efforts as (a) enhanced education about participation, (b) broader access to a variety of communication channels, (c) further development of communication skills, (d) widespread involvement in decision making, and (e) the development of genuine dialogue between and among various groups of organizational stakeholders.
Ultimately, then, workplace or organizational democracy should be understood in terms of a wide range of communication practices as well as with respect to economic control. And, given that these very practices may be seen as constituting democracy, they must be open-ended, adaptable, and subject to scrutiny, within any work or organizational context we may consider.
Rather than offering a detailed summary of this large review, in this concluding section we would first like to suggest a few areas for further exploration, in terms of theorizing, empirical investigation, discussion, and democratic practice. We return here to the categories that organize our review. First, it is clear that more work is needed concerning the relations between organizational democracy and wider practices in the community. In fact, we are led to consider at the broadest level the relationship between democratic practice and multinational corporate capitalism in an era of "globalization." This issue is of enormous practical as well as theoretical importance, particularly as numerous societies are considering ways to stimulate citizen participation and revitalize democracy. Specifically, the limitations of the market as a substitute for vibrant and expansive democratic practice should be examined closely. Also, the possibilities and the challenges of new computer-mediated technologies should be considered, especially in terms of their capacities for simultaneously encouraging some forms of participation and stifling others.
Second, the rationalities of and motivations for employee participation and workplace democracy need to be considered seriously in an age when the "social contract" between the individual and the employing organization is being questioned, redefined, and reshaped. We take the position that, although
the kind and extent of organizational democratization need to be geared to the nature of the work, all types of organizations would benefit ultimately by making genuine commitments to their employees and their goals (even though such commitments may be less than permanent). From a communication standpoint, the discourse surrounding the implementation of programs of workplace democracy and employee participation is revealing in terms of the motives and organizing symbols that come to predominate in any particular case.
Third, the possibilities for the survival of "postbureaucratic," relatively egalitarian organizations remain debatable. Although the evidence weighs heavily against the long-term maintenance of the "integrity" of highly democratic organizations, what remains to be assessed in depth are highly adaptable, process-oriented models of organization. Clearly, communication patterns within and between presumably democratic organizations and their environments are among the deciding factors in this issue.
Fourth, with respect to the "microprocess" or specific behavioral aspects of workplace democracy, it is important that the extensive knowledge gained from a century of leadership study be brought into the examination of contemporary programs of employee participation. The demands and dilemmas of team leadership must be explored more fully; behaviors and persuasive strategies should be examined in varied social contexts. At the same time, the prevailing and marginal discourses of the organization should be examined more closely for how the activities of leadership and participation are understood, framed, and practiced within them.
Fifth, the issue of employee voice should be investigated more thoroughly, both with respect to apparent opportunities for individual expression and in terms of organizational constraint. For example, it has become evident in recent years that a degree of coercion persists in organizations beyond what had been commonly recognized. In analyzing both unobtrusive and overt forms of control, we must consider that the line between them is often blurred, and that there can be covert forms of direction or even coercion (hidden from the observer's view), just as there can be not-so-subtle instances of unobtrusive control (as in deep and vocal allegiance to a corporate mission).
Sixth, the question of adversarial versus consensus-based models of organizational democracy represents an important practical matter for policy makers and decision-makers in organizations. Specifically, size/growth, value homogeneity, and means to achieving group cohesion all should be seen as topics for periodic review. Moreover, from a broader practical and philosophical perspective, our very notions of consensus have been challenged by postmodern critiques (just as presumptions about conflict have been challenged by feminist authors). This further intensifies the need for more detailed empirical analysis of the (non)expression of interests in real situations.
Finally, the ironies of power and control in presumably democratic organizations merit attention as well as self-reflection. Control in this sense must be seen simultaneously as inter-personal and systemic and as an immanent but changing dimension of interaction at work. We should therefore be attuned to the surprises, contradictions, and paradoxes entailed in organizational transformations toward presumably greater employee "empowerment," seeing these challenges as part of the ongoing processes of democratization and organizational change.
Despite our goal of comprehensiveness for this review, we must acknowledge its boundaries. First, for the most part, we have had to exclude consideration of other relevant literatures, notably those on new technologies, "diversity" in organizations, and social movements. Second, we have not been able to offer much in the way of historical contexts for the types of organizations and activities we have described and interpreted. Third, despite our intention to "internationalize" the review, we admit to a lack of awareness of much research that no doubt exists about organizations beyond Europe, North America, Russia, Japan, Australia, and New Zealand. We trust that future such reviews will be even more expanded in scope, reflecting both the importance and the dynamism of the theme of democracy, participation, and work.