Life is organized in rhythms, boundaries, and containers — organic patterns somewhere between formal mechanism and germinal open space. In the living laboratory of our on-line learning community, we use the groupware tools of open space, timing, rhythms, boundaries, containers, and procedures in changing patterns to support creative group activities.

We tailor our groupware to support the purpose and flow of our workshops, creating dynamic forms that embody evolving group purpose. Similarly, our workshops encourage participants to shape or tailor their lives according to what matters most to them — their personal purpose and values.

We believe that purpose-centered groupware will play an essential part in our transition to a vital, sustainable culture.

In 1978 we coined the term groupware, now a buzzword in corporate and academic circles for software that supports group work. Our original definition was broader: “intentional group processes plus software to support them.” We now realize that groupware is computer-mediated culture. Some parts are embodied in software, other parts in the hearts and minds of those using it. It is most effective when the software is tailored in support of a group’s purpose.

While the field of computer-supported cooperative work or groupware has grown in recent years, it is still in its infancy. Many approaches focus primarily on the technological part. Few systems are truly tailorable. Little is known about asynchronous group dynamics and the appropriate fundamental building blocks, or primitives, of asynchronous group process.

In our Conceptual Framework (Part I), we review two prevailing approaches to groupware. Each is based on a causal principle reflecting a particular world view: (1) groupware as mechanism that makes groups work (right-handed causality, world as machine), and (2) groupware as context or open space that allows groups to self-organize (left-handed causality, world as matrix). While each of these plays an important part, neither is sufficient by itself.

As our global culture moves beyond the hierarchical social systems which served a vital role in our evolution, the old social forms which once gave meaning to and provided guidance for our lives are changing. Our institutional, community, and family structures appear to be crumbling beneath us, and we are tempted to grasp for something to hold onto. This is the natural right-handed impulse to create forms to hold life in. But when this fails, when the turbulent flow cannot be held, we are tempted to the other extreme, to let go of forms. This is the natural left-handed impulse to “create” life by letting it flow. But that doesn’t work by itself either. Our social fabric unravels in the unlimited “freedom” of no commitments, no rhythms, no patternings.

Rather than replacing the passing hierarchical forms with computer-supported versions (groupware as mechanism) or providing minimal social forms (groupware as context), we are exploring ways of using both right and left hands together. By seeing group life as a living system and looking for ways the machine can embody organic processes which support that living system, we can begin to integrate groupware as context and groupware as mechanism.

Based largely on Charles Johnston’s theory of creative causality (Johnston, 1986), our conceptual framework builds on an emerging world view which bridges right- and left-handed causality (world as creation). Form and context are part of a dynamic creative process: form (software) continually emerges from and evolves in response to its living context (group life).

We summarize three powerful conceptual tools adapted from Johnston’s theory and elaborate on them in the Appendices. His creative cycle (Appendix B) differentiates any creative process into distinct stages with unique dynamics which inform our model of the stages of asynchronous group evolution (Appendix E). His approach to bridging polarities provides a way of integrating creatively the polar opposites that mark the core issues in groupware, such as context-form and human-computer (Appendix C). His referents of aliveness and capacitance (capacity to create) offer more inclusive alternatives for measuring the effectiveness of groupware than productivity (Appendix D).

We also adapt aspects of Arnold Mindell’s process oriented psychology (Mindell, 1985) by shifting our fundamental units of analysis from software to living processes which emerge from the background into the foreground and primary form of the software (Appendix A).

Using this conceptual framework, we develop a third approach to groupware as creative process that reflects this new world view. Effective groupware can be tailored with an array of group process primitives adapted from Johnston’s (1990) “bag of tools” for change agents. These organizing tools are the foundation of all living systems. They bridge context and mechanism in six functionally distinct steps: white space, timing, rhythms, boundaries, containers, and procedures.

In our Action Research (Part II) we use these primitives to support our on-line learning community and self-development workshops. Our educational approach is purpose-centered — helping participants discover what matters to them (their purpose) and shape their lives from it. This creative process is similar to our approach to groupware which begins with discovering what matters to a group (its purpose) and then shaping the software to support it.

In the living laboratory of our learning community, we are exploring ways of providing flexible, tailorable, temporary forms, particularly at the early stages of group development. These “training wheels” guide groups into coherent patterns, persuade them into making and keeping commitments, and nudge them into creative interactions by providing rhythms, boundaries, and containers that hold and support group life — but not too tightly!

Our tailorable groupware toolchest includes scripts as agents for group facilitators, several different communications arrangements for group work, and other tools for implementing the primitives. Our primary rhythm tool, the circle arrangement, uses spatial and temporal boundaries to create rhythmic containers for group life, persuading “lurkers” to participate and nudging the group into a rhythmic “heartbeat” of activity. Our participation data demonstrate the effect of these rhythms, boundaries, and containers and support the utility of our approach.

In Part III, we conclude by listing principles of purpose-centered groupware that (1) evokes what matters to a group (its purpose), and (2) augments its capacity to create forms which embody that purpose. This requires tailorable software and human facilitation, both of which are most effective when they persuade rather than manipulate a group into making wise choices. Integrating our educational and technological methods into a creative whole, we believe such groupware applies potentially to any computer-supported group work. As the virtual reflection of a learning society, it is an essential part of the foundation for our necessary transition to a vital, sustainable culture.

We have been students of groupware* ever since we coined the term in 1978 (Johnson-Lenz, 1980, 1981a, 1982). We expect to be its students for decades to come. While Fortune magazine (Richman, 1987), Robert Johansen’s book on groupware for business teams (Johansen, 1988), and many recent articles have made the term popular, groupware is still in its infancy. Seen in the context of the long journey ahead, even the most sophisticated approaches today are but baby steps into the uncharted frontier of hyperspace — that asynchronous, aspatial place where we meet via computer (Johnson-Lenz, 1989c).

We believe the wise inhabitation of this frontier is a key to our future. Technologists and futurists predict we will be doing much of our business, education, and personal interaction via computer in the next century (Nilles, Carlson, Gray & Hanneman, 1976; Hiltz & Turoff, 1978; Johansen, Vallee & Spangler, 1979; Toffler, 1980; Hiltz, 1984; Wright, 1990). Futurists and social theorists also predict a necessary and fundamental restructuring of society in response to global environmental crises (Fuller, 1969; Meadows, Meadows, Randers & Behrens, 1972; Schumacher, 1973; Roszak, 1979; Henderson, 1981; Hawken, Ogilvy & Schwartz, 1982; Naisbitt, 1982; Markley & Harman, 1982; Harman, 1988; Naisbitt & Aburdene, 1990; Theobald, 1990). The speed and complexity of the changes already beginning to occur are unprecedented in human history. We believe appropriate and flexible groupware is an essential part of our capacity to respond effectively to these changes.

What is effective groupware for these times? This overarching question is the focus of our inquiry.

Groupware is computer-mediated culture. It is an embodiment of social organization in hyperspace. Because many of the global problems we are facing are widely recognized as symptoms of limitations inherent in the older mechanistic world view (Bateson, 1972, 1979; Berman, 1981; among others), groupware which serves in these turbulent times will necessarily reflect more sustainable cultural patterns than those which no longer work.

As culture, groupware is a single system integrating co-evolving human and tool systems. Our original definition of groupware is “intentional group processes plus software to support them.” It has both computer and human components: software in the computer and “software” in the people using it. The human component is a shared “mental model” of what the group is doing (its purpose) and how it is doing it (its process). We have recently extended this definition to include other more expressly cultural factors, including myth, values, and norms (Johnson-Lenz, 1989a). The computer software should reflect and support a group’s purpose, process, and culture. Like well-designed co-processes in a cooperating network server and workstation, the matched co-processes of computer and group software join the machine and biological “hardware” in a creative whole.

Engelbart recognized the power of this co-evolutionary approach and proposed a “bootstrapping strategy” whereby each generation of groupware technology is used as the creative context within which the next generation is designed and implemented. “Inherent in the Bootstrapping Strategy is the explicit co-evolution of the Human-System and the Tool-System as one holistic system. As the complexity and urgency facing today’s organizations increases exponentially, the need for this Co-Evolution approach becomes absolutely critical.” (Engelbart, 1988, p. 5)

TAILORING GROUPWARE TO FIT GROUP PURPOSE

In times ahead we should find ourselves more able to understand institutional structure itself not as truth, but as a context for creative process....In other words, we should find people increasingly adapting structures to fit their purposes, rather than tailoring their lives to fit structures. (Johnston, 1988, p. 10)

It is possible to design, via the computer, almost any variation from a free and open communication process to a complete dictatorship....The list of possibilities for imposing control are rather large....In that sense, this technology is a rather unique two edged sword. (Turoff, 1982, p. 255)

In 1977 we began working with Murray Turoff and Roxanne Hiltz via the Electronic Information Exchange System (EIES) housed at the New Jersey Institute of Technology. From Oregon, we collaborated on research and development entirely via computer before meeting Turoff in person in 1979 and Hiltz in 1980.

Turoff’s uniquely tailorable toolchest and concept of “structured communication” (Turoff, 1971; Hiltz & Turoff, 1978) drew us to EIES. From our early experiments there we learned two fundamental lessons: (1) some structure is needed by most on-line groups, and (2) different structures serve different purposes. We began to realize the software component of groupware is best tailored in response to the needs of the human component, rather than the other way around. The direction of the arrow in Figure 1 is critical if groupware is to effectively evoke and support the full measure of our creative capacity. While this may be obvious, it is neither widely understood nor practiced.

Values and cultural assumptions are incorporated into each and every groupware design. Often they are not readily apparent (Appendix F: Hidden Assumptions in Groupware). Designers may be wise enough to create structures which empower. That may be a necessary first step in the “bootstrapping” process, but a better solution is to involve users in the design process. Interest in this approach is growing as indicated by the recent Participatory Design Conference, PDC ’90, sponsored by Computer Professionals for Social Responsibility.

Better yet is a design that allows its users to pattern their own communications by providing tools for those directly involved (facilitators, managers, and even users) to tailor the system to their purposes. The effectiveness of such tailorability depends on identifying the appropriate primitives of human interaction via computer.

But there is more to it than the right primitives. Just as designers need to learn how to design tailorable toolchests, groups need to learn how to choose the structures they want to inhabit in hyperspace, even in a highly tailorable system. Figure 2 is an adaptation of a diagram of the groupware design process from our early work (Johnson-Lenz, 1981b).

This is the two-fold goal toward which we reach in our work: (1) groupware which is tailorable by those using it, and (2) group processes that help groups develop the maturity to determine their own course and to use the technology to support their purposes.

In the remainder of Part I, we present our conceptual framework for groupware. We begin by reviewing two prevailing approaches: groupware as mechanism and groupware as context. Each of these holds a part of the answer to our overarching question of what is effective groupware. In Part II we describe a new class of tailoring primitives and our experience using them.

GROUPWARE AS MECHANISM

Groupware as mechanism is the most common approach to groupware. It is based on the social theory that human interaction can be modeled as a machine with deterministic interactions among separate parts. Examples include hierarchical chains of command and Robert’s Rules of Order. This social theory is a reflection of the mechanistic world view that rose into dominance during the Industrial Revolution. If the universe is a machine, so is society.

From this point of view, the computer’s role in groupware is to provide well-defined mechanisms for interaction. Much of the groupware currently available in the corporate marketplace falls into this category, including many computer-supported cooperative work systems. The most common examples are systems for shared authoring, shared calendars, and project management that regulate interactions within a limited, task-specific focus.

More ambitious designs seek to encompass a broader range of tasks with a general-purpose mechanism. Winograd and Flores’ The Coordinator (Winograd, 1986) is the first of a genre of systems that define specific communication acts — requests for action and commitments to fulfill them. They require all interaction to fit within that framework. The computer monitors the fulfillment of all commitments. Transaction groupware, as defined by Esther Dyson (1989), breaks complex tasks into replicable transactions. Its purpose is to improve efficiency by making all transactions explicit and automating the routine parts.

Social coherence in these systems is “caused” by a mechanism that determines the outcome through explicit rules and algorithms. The causal principle is mechanistic, or what we call right-handed causality, popularly called “cause and effect.”

Such groupware designs focus on the basic units of interaction among people cooperating on well-specified tasks. They are designed to keep the group on task, enforce roles and commitments, make the group efficient and productive, and make it behave in a certain predefined way.

The implicit values often underlying such groupware are control and adherence to rules, at the expense of individual autonomy and freedom of choice. Often their inflexible structures trigger organizational and individual resistance, effectively reducing the creative capacity of the group rather than increasing it. Zuboff (1988) criticizes the use of computer technology to monitor workers, warning of the potential danger of electronic sweatshops. Reder & Schwab write that:

Automated attempts to “pin people down” and thereby enhance accountability may not bring about better communication or enhanced productivity. It is very likely such attempts, if they are accepted by users, will change the “rules of the game,” and certain types of critical conversations will move to contexts in which the tool will not be used, thereby altering the nature of the communication which does take place through the new technology. (Reder & Schwab, 1988, p. 367)

While these criticisms are sometimes warranted, it is most often when the groupware has taken the mechanistic model to the extreme. There is an important and useful place for groupware as mechanism. It is part of the answer to our question of what is effective groupware, but it is not the whole answer. Something more is needed. The critics point the way. More flexible patterns that encourage personal initiative are just as important as well-defined group procedures.

We have oversimplified our description of these systems in order to highlight the dominant causal principle in their designs. Most include some flexibility. For example, recent versions of The Coordinator have added more flexible communication act categories. This only serves to make our point.

GROUPWARE AS CONTEXT

Groupware as context is the popular “alternative” approach to groupware. It is based on the social theory that human systems are self-organizing and arise out of the unrestricted interaction of autonomous individuals. This social theory has emerged in response to the limitations of hierarchical, mechanistic styles of organization and is part of a much broader challenge to the mechanistic world view in virtually every field.

Much has been written about this social theory applied to management. Several works are particularly relevant here. Linda Ackerman (1984) has developed a theory of “flow state management” in which the primary management task is removing blocks to the natural flow of energy. Tom Peters (1987) describes the advantages of actively engaging with chaos as a creative source rather than trying to control it. Harrison Owen (1987) writes about creating “open spaces” with few, if any, forms to pattern interactions and within which people organize themselves.

From this point of view the computer’s role is to provide open space for people to meet and self-organize. Most conferencing systems such as CompuServe, The WELL, the Santa Monica Public Electronic Network (PEN), and America Online embody this approach with variations on a simple two-tiered structure originally used on the Confer system (Zinn, Parnes & Hench, 1976) that allows participants to organize themselves into topics or special interest groups.

In some of the more sophisticated systems, the computer also provides tools to support this self-organizing activity. Incorporating precursors to Sculley’s (1987) vision of a knowledge navigator, two levels are usually involved: (1) collective knowledge structuring tools, and (2) personal interface tools for navigating those structures. Knowledge structures provide forms within which communications can be organized into more useful patterns using hierarchical file systems, keywords, and hypertext. User interfaces support pull-down menus and point-and-click hypertext links.

Engelbart’s (1962) Augment pioneered most of these concepts. Newer commercial systems including Connect and America Online incorporate some of them. The recently released Lotus Notes system is a derivative of the Xerox NoteCards system (Trigg, Suchman & Halasz, 1986) which extends many of these ideas. Our system (Johnson-Lenz, 1988a) includes simple hypertext tools similar to those in the original Augment system. Turoff’s EIES2 goes a step further with group and personal organizing tools using a “list processing“ metaphor (Turoff & Hiltz, et al., 1989). Even low-end conferencing and bulletin board systems such as Coconet, HyperBBS, and TeamMate are beginning to reflect some of this sophistication by including hypertext capability and point-and-click interfaces.

The causal principle of social formation in these systems is non-mechanistic, or what we call left-handed causality. It assumes that social coherence will emerge from individual interactions. It is the polar opposite of mechanistic causality which seeks to produce coherence. Left-handed causality “causes” coherence by creating space for the coherence to emerge and then allows it to happen.

At the group level, this is non-mechanistic. However, it is mechanistic at the levels of knowledge organization and user interface. Groups form as the result of individuals using personal knowledge navigation tools interacting through a vast knowledge building mechanism.

These systems focus on the user interface and structure of knowledge. They are designed to give users as much freedom as possible. Rather than focusing on commitments and support for keeping them, such systems offer the polar opposite — complete freedom to come and go at will, to do whatever one wants, whenever one wants to do it.

The implicit values embedded in these systems are individual autonomy and freedom of choice at the expense of commitment and responsibility which are necessary for social coherence. This is not really groupware. If groups do emerge, it happens without much support from the software. The results are often disappointing, or at least sporadic. Sometimes there is a lot of activity, but conversations tend to be scattered and of uneven quality. Small islands of creative interaction do occur, but they are infrequent (Johnson-Lenz, 1989b, 1989c).

This approach is an important alternative to the mechanistic approach, but it either reduces social interaction to chance meetings or to the mechanical accumulation of knowledge. The embodiment of culture as groupware in such systems is still mechanistic.

This open space approach offers another part of the answer to our question of what is effective groupware for these times, but it is still only a part, even when taken together with the mechanistic approach. At that they only scratch the surface. Routine transactions and explicit commitments can lack meaning. A user interface is but a single membrane in a rich and complex living social system. What and where is the deep structure?

VIRTUAL FORM IN HYPERSPACE

Since we’ve seen each other, a game goes on.
Secretly I move, and you respond.
You’re winning, you think it’s funny.

But look up from the board now, look how
I’ve brought in furniture to this invisible place,
so we can live here.

(Rumi, 1984 trans., quatrain 1245)

Within hyperspace, individuals and groups create virtual forms — objects in essence and effect, but not physical form. The simplest contain information — text, graphics, or sound. Of greater interest here are groupware objects which contain groups and group procedures, giving form and meaning to individual interactions within them. A variety of virtual social forms have been written about recently, including the Virtual Classroom™ (Hiltz, 1986), a Virtual Library (Wheeler, 1987), virtual communities (Rheingold, 1987), strolls through virtual neighborhoods (Jacobson, 1989), and a virtual Cyberspace Playhouse (Walser, 1990).

Both groupware as mechanism and groupware as context contribute essential parts to the process through which virtual forms emerge in hyperspace. Groupware as context provides open space. Group work begins with the empty context of hyperspace — a container for creative interaction and collective knowledge building. Knowledge structuring software is necessary. Groupware as context also brings the creative capacity of individuals to that context. Group work arises out of their interaction. User interface and knowledge navigation software is necessary. Groupware as mechanism provides group procedures. Coherent group work requires commitment, well-organized transactions, lists of tasks, and more. Group procedural software is also necessary. Table 1 summarizes the characteristics of both approaches.

Each of these plays an important part. But something important is still missing. What fills the forms with meaning? What deep structure behind the user interface connects these parts into a creative whole?

THE VITAL MISSING LINK

What is missing? There are many ways it could be put, but perhaps most simply, what is missing is life. In Newtonian/Cartesian reality the universe and all within it are like a great clockworks, an immense and wondrous piece of machinery. A mechanical paradigm can offer us many amazing things. But no matter how great a machine’s complexity, it can never be more than just that, a machine. (Johnston, 1986, p. 7)

A social theory that bridges the mechanistic and contextual theories discussed so far is beginning to emerge. It reflects a new world view broad enough to include both right- and left-handed causality. Instead of rejecting mechanism, it integrates it into a larger framework. Rather than viewing reality as a machine, a rule-driven, deterministic system of fixed, independent parts, it views reality as a creative process, essentially interconnected, multiply determined, inherently uncertain, and constantly changing (Johnston, 1990). Rather than measuring the effectiveness of a system in mechanistic terms alone, such as productivity or efficiency, it uses more inclusive measures or referents such as its capacity to be creative, to be alive.

This integral world view and the social theory that reflects it are the foundation of our conceptual framework. They provide a way of thinking about group process via computer which encompasses the need for mechanisms to support group work as well as the need for open spaces in which group life can emerge. They provide powerful conceptual tools for integrating these two approaches into a single, creative whole. Two bodies of work have been particularly influential in the development of our framework: Mindell’s process oriented psychology (Mindell, 1985) and Johnston’s creative causality (Johnston, 1986). We summarize these works briefly below as background for what follows. We describe them and their application to our conceptual framework in greater detail in Appendices A through E.

Mindell describes individuals and groups as having foreground and background processes, or what he calls primary and secondary processes. A primary process is what appears to be happening at the moment. A secondary process is what is emerging to challenge the primary process, the growing edge of an individual’s or group’s life — the vague dissatisfactions, inchoate preferences, or senses of what could be better. His work provides two important reframes for groupware: (1) individual challenges to group form are a source of creative power for rather than a threat to a group, and (2) the fundamental unit is a process rather than the channel (primary/foreground, secondary/background) in which it operates. Groupware is more than explicit form or empty context. It is the living, creative process of evolving form in hyperspace. For more detail, see Appendix A: Primary and Secondary Processes.

Johnston’s work provides us with even more powerful tools for our conceptual framework. In his second book, Johnston (1990) shows how our tendency to polarize issues as “either/or’s” limits our thinking. Bridging polarities involves framing questions in more creative “both-and” ways. Effective groupware necessarily bridges many polarities including human-machine, context-form, and autonomy-commitment. It is a creative, dynamic balance of polar opposites. The appropriate balance depends on the group, its stage of development, and its purpose. For more detail, see Appendix C: Bridging Groupware Polarities.

The referent by which we measure our work often determines whether we polarize issues or bridge polarities. Measures which reflect the creative, dynamic nature of living systems are most appropriate for bridging. Johnston suggests two such integral referents: capacitance, which he defines as the capacity to create, and aliveness, which is intuitively obvious, although subtle. For measuring the effectiveness of groupware, we prefer these to productivity or the alternative, sustainability. Effective group work involves production and sustenance. Capacitance includes both the capacity to produce and sustain. Aliveness is more than mere activity. Rest and reflection are vitally important to the creative process. For more detail, see Appendix D: Post-Mechanistic Referents for Evaluating Groupware.

Johnston’s theory of creative causality bridges left- and right-handed causality with “world as creation.” His creative cycle divides the creative process into stages through which form emerges and differentiates from its creative context. This creative process applies equally well to an individual life, a relationship, a creative project, and global culture. Applied to groupware, form emerges from group life in distinct stages with unique organizing dynamics. Different groupware primitives are appropriate at each stage. For more detail, see Appendix C: The Creative Cycle and Appendix E: Stages of Asynchronous Group Evolution.

GROUPWARE AS CREATIVE PROCESS

If a wave only sees its form, with its beginning and end, it will be afraid of birth and death. But if the wave sees that it is water, identifies itself with the water, then it will be emancipated from birth and death. Each wave is born and is going to die, but the water is free from birth and death. (Thich Nhat Hanh)

One of the paradoxes of group life is that a primary process is necessary for any shared activity but it also dominates the group’s attention. A shared form to focus the energy is vital. Without it nothing would happen. But that form is an output of the living process, not the process itself. Temporarily forgetting the context from which it arose is an essential part of the creative cycle, but it is not the whole cycle. To be effective, at times groupware must hold the group life and at others get out of the way so that attention can turn to emerging secondary processes and new forms appropriate for their support can be created.

Effective groupware must have the capacity to support (1) both primary and secondary processes, and (2) both the differentiation and integration phases of the creative cycle. In face-to-face groups secondary processes appear in background channels like body talk where discomfort or boredom can be expressed without interfering with the group. Via computer, background channels like electronic mail and places for debriefing or evaluating group work provide a similar function. Reder & Schwab (1988) have shown that users spontaneously select different channels for different communication purposes. But isolating secondary processes in other channels still does not provide a way for them to emerge into primary form within the group. Other tools are needed, tools which respectfully invite the expression of secondary processes within the group and which support them in influencing the direction of the group.

GROUPWARE AS RHYTHMS, BOUNDARIES, AND CONTAINERS

The problem of making computers useful to people as communications and information devices is not an engineering problem, it’s a design problem. Engineers are trained to eliminate the subjective factors. But it’s exactly the subjective factors that are critical here. (Kapor, 1990, p. 85)

All living systems are rhythmic. Everything turns in circles within circles. The daily rhythm of sunrise and sunset is contained within the broader cycle of the seasons. The rhythms of our heartbeats and breathing give us life.

Living systems are contained in vessels with flexible, permeable boundaries. The biosphere contains all life on the planet. Our bodies contain our organs which contain cells which contain cell nuclei and so on.

Rhythms, boundaries, and containers are primitives — universal, fundamental patterns from which all life is built — including our social life. Our face-to-face contacts often occur in regular rhythms. Boundaries of many sorts pattern when and where we connect and when we don’t. Physical and social containers frame and hold our meetings. The skillful use of these tools is second nature to experienced group facilitators. Might groupware based on these primitives more effectively support creative group life?

Rhythms, boundaries, and containers are patterns that connect groupware mechanisms and germinal open spaces in a creative whole — fundamental dynamic patterns to hold group life in hyperspace, and on whose foundation other forms are built. Figure 5 diagrams this creative process.

In the fall of 1988, after a year of prototyping and pilot testing, we convened our Virtual Learning Community for self-development education. It uses a computer-mediated communications system and tailorable groupware of our own design. In this community, where we are active participants, we are exploring appropriate rhythms, boundaries, and containers for creative group work via computer. While our focus is self-development education, we believe these primitives apply to almost any asynchronous group work.

Members connect to our UNIX host via modem at 1200 or 2400 baud. Because our members have a wide variety of personal computing equipment, our interface uses the popular VT100 display standard that supports visual editing and simple graphics. More than half of our current members are outside the Portland, Oregon metropolitan area. The system is available 24 hours a day, 7 days a week. Activities are asynchronous; members participate at times of their choosing.

Workshops are the backbone of our learning community. Each focuses on some aspect of the creative process of living, or what we call living on purpose. Our curriculum applies the principles of creative causality to self-development. It is an example of what Johnston calls purpose-centered education: “to help people be in touch with what matters most to them, what for them is most fully alive, and find the courage and capacity to shape their lives from it.” (Johnston, 1988, p. 6) Living on purpose is isomorphic to groupware as creative process. Group life on purpose is groupware as creative process in hyperspace. Note the structural similarity between Figure 6, The Creative Process of Living on Purpose, and Figure 2, Groupware Tailoring Process.

Education in this sense goes far beyond personal growth workshops. Based on our experience we believe that incorporating self-development principles into groupware often increases the effectiveness of on-line groups and aligns the group’s work with the cultural shifts which are central to our collective survival (Johnson-Lenz, 1989b). While beyond the scope of this paper, we expand on this hypothesis briefly in our concluding remarks.

Members of our learning community participate in an evolving variety of on-line activities. Rather than listing them all, we describe our experience with three selected activities:

• Right Livelihood Workshop, September to December, 1988: a 12-week workshop with 14 participants looking for more meaning and fulfillment in their work.
• The Creative Exchange, November 1989 through April 1990: an open-space activity with 26 participants who self-organize into small groups to explore common personal and work issues.
• Living on Purpose Workshop, March 1990: a 4-week workshop with 14 participants exploring their personal life purposes.

To support our learning community and action research, we have developed our own tailorable groupware. Written in C and running under UNIX, it includes:

• items: text objects
• exchanges: named containers for exchanging items which are organized in a hierarchy of sub-exchanges with levels separated by periods; examples: right livelihood.circle, creative.talk, purpose.1.sharing
• actions: commands for performing operations on items and exchanges; examples: view <item>, select <exchange>, download new <exchange>
• permissions: on/off switches for an exchange, permitting specific classes of actions within it; examples: view items, change own items only, create sub-exchanges
• roles: named roles for an exchange with associated permissions; examples: convener, facilitator, member
• members: users permitted by role to perform actions within an exchange
• markers: pointers for a member to last item viewed, created, or responded to
• group markers: pointers for an exchange marking a range of currently active items
• settings: on/off switches for an exchange, specifying structures and governing interactions within it; examples: private membership, pen named/anonymous items allowed, items named rather than numbered
• arrangements: named, commonly used patterns of roles and settings; examples: meeting, topics, circle
• scripts: items containing procedures written in a structured, high-level interpreted programming language which includes actions and variables for accessing the above parameters
• menus: items containing menu text and commands that map abbreviations (menu choices) to specific actions and which adapt the displayed menu text to the user’s situation (intelligent menus)
• agents: scripts that act on behalf of facilitators, either at pre-determined times or in response to certain events

ARRANGEMENTS

Just as the arrangement of furniture and other objects in a room has a significant effect on how people interact, the arrangement of objects in hyperspace governs our interactions there. Benston (1990) notes how software “furnishings” set the stage for social interaction via computer. Successful facilitators pay careful attention to how the space is arranged so that it fits and supports the group’s process needs. They need tools that allow them to rearrange the furniture.

Unlike most systems which provide only a single arrangement, our toolchest provides a variety of arrangements, some which emulate the most common structures available on other systems, and some which are unique to our system. Spaces can be rearranged by facilitators — without reprogramming. The three arrangements of particular importance here are:

• meeting: an exchange of sequentially numbered text items; the system keeps a marker for each member showing the last item viewed; users may enter any number of items whenever they want; the simplest form of asynchronous conferencing
• topics: an exchange of sequentially numbered text items called topics plus a sub-exchange for each with sequentially numbered responses; the system keeps a marker for the last topic viewed and the last response viewed within each topic; users may enter new topics or responses whenever they want; two-level conferencing
• circle: an exchange of sequentially numbered text items called rounds plus a sub-exchange for each with responses identified by user name; the system keeps a marker for the last round responded to and the last response viewed within each round; users may enter only one response to each round and may respond only to certain rounds at any given time

POST-MECHANISTIC GROUPWARE PRIMITIVES

Information is a derivative of the relationship, not the other way around....Instead of asking, “What is the information that matters and how do we most effectively manage it?” companies must start asking, “What are the relationships that matter and how can the technology most effectively support them?” (Schrage, 1990)

Using the above tools for building blocks, we have implemented six groupware primitives adapted from Johnston’s (1990) “bag of tools” for change agents. We call these post-mechanistic since they bridge groupware as context and groupware as mechanism. These primitives are:

• white space (context)
• timing
• rhythms
• boundaries
• containers
• procedures (mechanism)

WHITE SPACE

The function of white space, or context, is creative incubation. It is the germinal source of form. Spatially it provides a context for creation to emerge. Temporally it provides time to reflect, to let the answer come.

The most effective amount of white space depends on the situation. If existing forms seem rigid and stuck, more space may allow new forms to emerge. If lively interactions are not occurring frequently enough, a smaller space may gather the energy together. If the group is grasping for answers, more time may allow new insights to emerge. If people are scattered, shorter timelines may focus their activity.

Our Creative Exchange is an example of open space. Using the topics arrangement, it offers a context within which members self-organize into mutual interest groups. However, it is not just open space. As detailed in later sections, we provide a variety of forms to get things going.

We provide temporal white space, or open time, by reminding members to take time to relax and reflect. Optional guided imagery and pauses in text to display selected words and phrases slowly and rhythmically offer “time medicine” for reducing stress and finding the calm center in the midst of busy, pressured lives (Dossey, 1982). Participants frequently report getting much more from our workshops and other activities when they follow such reminders and take time to allow their creative process to unfold (Johnson-Lenz, 1988b).

White space and containers are related. White space is created by an empty container. Every container in our system begins as white space.

TIMING

The functions of timing are threefold: knowing when to use what tools; punctuating transitions, particularly beginnings and endings; and shifting to the next stage of group development.

Right timing is a dynamic balance between sooner or later. When tools are used at the right time, they increase the capacitance and aliveness of the group. When used too early they get in the way. When used too late, they have little effect.

While knowing when to use them is an art we are still learning, we have discovered a few useful guidelines:

• More form is needed during the early stages of group evolution (Appendix E). Just as a parent guides a child’s development and then gets out of the way as it matures, the trick is knowing when to let go.
• Timing and boundaries are related. All containers have beginning and ending boundaries. Timely use of transitional groupware forms can help a group come together at the beginning and complete at the end.

• An orientation software agent requires members to view guidelines and complete other one-time new member tasks before participating.
• Most activities have opening and closing ceremonies using the circle arrangement.
• We often begin a temporary process for participatory evaluation or negotiation of next steps near the end of activities.

RHYTHMS

The function of rhythms is to provide appropriate patterns for periodic contact and participation. Most face-to-face meetings occur in rhythms that we take for granted, whether daily, weekly, or even yearly. Gibb (1978) has written that rhythm and flow are essential for creating high-trust groups. Rhythm is an embodiment of commitment to relationship.

Effective rhythms are a dynamic balance between slow and fast. If members do not feel connected with each other, a faster rhythm may bring them closer. If too much is happening, a slower rhythm may work better. Effective rhythms are both a tonic and an antidote for overload.

Appropriate rhythms are also a dynamic balance between fixed and permeable boundaries. If members are not connecting effectively, more clearly defined rhythms may help. If members are resisting the rhythms, more room for exceptions may allow greater energy to flow through the group. It is often challenging to find the right balance between individuals’ rhythms which may vary widely. The asynchronous medium makes this easier, but some common pattern is still necessary for a group to work. What seems to work best is a balance of reminding and persuading group members to participate regularly within a clearly-defined window and to allow some flexibility. See “Iterative Development of Circle Management Tools” below.

BOUNDARIES

The functions of boundaries include defining group membership; delineating group identity; and marking group rhythms, beginnings, and endings.

Living boundaries are inherently dynamic, varying along the dimensions of permeability and flexibility. A permeable boundary is one that allows things to flow through it. A flexible boundary is one that changes over time. This applies to both spatial and temporal boundaries.

An appropriately permeable boundary retains the identity of the group while allowing enough flow in and out to keep it alive. If a boundary is too permeable, it’s too easy to join or leave a group, or if the beginning or ending is unclear, energy may leak away. If a boundary is not permeable enough, if joining or leaving is difficult, or if beginnings and endings are too rigid, some individuals may resist or break the boundaries anyway. The right balance depends on the situation.

Flexible boundaries allow the primary process of the group’s current form to change in response to its emerging secondary process, its evolving “shape.” If a boundary is too flexible, if facilitators don’t limit the direction of the group, or if milestones are not enforced, its goals may not be reached. If a boundary is too rigid, there is no way of accommodating diversity and new directions within the group, or if milestones are immovable, group evolution ceases.

Most asynchronous conferencing systems have lurkers, participants who read but never express themselves, who “lurk” in the background. This common phenomenon exemplifies the situation-dependent balances necessary for effective group boundaries. In open spaces such as our Creative Exchange, where groups are self-organizing and their boundaries are highly permeable, there will be lurkers. If the purpose is to allow self-forming groups to emerge, some lurking is inevitable and not a problem. In contrast, if the purpose is to involve everyone, lurking is a signal of leaky boundaries. Equal participation is expected in our workshop circles. Group boundaries are clear. Commitments to participate are required. However, the price of full involvement is resistance.

Boundaries always create resistance on both sides. Energy on the inside pushes out; energy on the outside pushes in. Resistance is a sign the boundaries are functioning in a living system. A rough gauge of whether they are appropriate is if the resistance is equal in both directions. Stronger resistance on one side may mean the boundary would work better if moved. While keeping the flow within clear boundaries is often appropriate, we are learning that semi-permeable, somewhat flexible boundaries that persuade rather than control are more effective. They allow some of the resisting energy to move through them or to change the shape of the form, but not too much, at least during the early stages of group evolution.

CONTAINERS

The function of containers is to hold the energy, life, identity, or “presence” of the group. They must be big enough to hold the group in the most creatively alive way, but not too big. Containers hold both the white space within which the group develops as well as the procedures which support that development.

An effective container is a dynamic balance between the forces holding the group together and pulling it apart. It embodies the group’s commitment and other holding forces while resisting the diversity of impulses that do not contribute to the group’s purpose. While it can remind and persuade members to participate, if the strength of their actual commitment is significantly less than what is embodied in the container, the container will generate more resistance than it can hold. The primary form will give way to the secondary process of the group coming apart. If the strength of commitment exceeds that of the container, the group will probably acknowledge it by creating a new container that embodies the emerging secondary process of deepening commitment.

As a group develops within the original container provided by facilitators, particularly as it moves beyond the third stage of asynchronous group evolution (Appendix E) into embracing the diversity of its members and using disagreement creatively, it will need a larger container. In the early stages of most groups, diversity and disagreement is avoided. Participants hold back parts of themselves which do not fit with the primary process of the group. But as they begin to express those parts in constructive ways within the group, the group’s capacity actually increases to hold more life. It becomes big enough to hold the disagreements in a more creative way.

Not only is the size of a container important, but most groups need a variety of spaces serving different purposes. The boundaries may differ in permeability and flexibility and the procedures within them will be tailored to the purpose of each container. For example, our Creative Exchange includes a topics arrangement container with permeable containers within it for self-organizing groups. It also uses a meeting arrangement container for bringing the whole group together. In contrast, our workshops use a circle arrangement container for a formal circle process with clear boundaries encompassing all members, plus a meeting arrangement container for informal sharing. Both are needed to support the purpose of the workshop which is served by the rhythmic, equal participation of the circle and the arrhythmic, self-selected participation of sharing. The circle carries the primary process of the group, and sharing is an opening for secondary processes to express themselves.

Boundaries and containers are related. A container is defined by the boundary which differentiates the inside from the outside. Each side of the boundary serves a different purpose. The outside protects the group from unwanted interaction. The inside nurtures the group into being. From the outside the boundary defines a form. From the inside it defines white space.

Rhythms and containers are related. Rhythms provide a regular pattern of cyclic containers for group energy. Our circles offer weekly containers (rounds) within the larger container of the circle arrangement. Each focuses group attention and energy on a particular question or issue. In our workshops we have found that weekly rounds work best. With two-week rounds, groups seem to do little during the first week. Shorter cycles are too fast. The greatest aliveness seems to emerge within a weekly rhythm.

PROCEDURES

While procedures are fundamental to computing, in our work they serve several functions directly related to our five other post-mechanistic groupware primitives. These include sequencing events and managing boundaries and rhythms. Most of our procedures are embodied in scripts and agents. Many of them provide functions similar to what is commonly thought of as groupware.

The dynamic balance central to effective use of procedures involves knowing when to use them and when to let the life of the group manage itself. Too much procedure and the group rebels. Too little and nothing happens. In our experience, they work best in collaboration with the natural life of the group, nudging, guiding, persuading, or reminding users rather than controlling them. For example, the design of our circle agent described in detail below has evolved from one which controls to one that persuades. Rather than forcing anything, it invites members into new spaces and reminds them of incomplete tasks which they can choose to do or leave incomplete.

On the other hand, we also enforce certain procedures without allowing any exceptions. For example, an orientation software agent assures that new members view the guidelines, view and agree with the community covenant, and complete their membership directory entry before participating in our workshops.

One of the growing edges of our work is continually observing the flow of group life and changing the sequence of events and paths that users follow. We are learning to support what is most alive for the group. This includes (1) the lively parts of primary foreground processes, (2) the incomplete parts of processes that are falling into the background, and (3) emerging secondary processes that are not yet dominant but may be vital in the future. It’s a challenging and often elusive balancing act. Sometimes we miss the mark. Rather than trying to control the group to fit our plans, we are learning to lead by serving its needs (Greenleaf, 1979), persuading it to follow the path that seems most vital.

TIMELY EMERGENCE OF VIRTUAL CONTAINERS

A useful technique involves the timely appearance of a virtual container directly in path of participants. When its purpose has been served it may disappear completely or move to the side of the path, out of the way. It’s like a special-purpose room appearing out of nowhere in the middle of a professional conference, just as it’s needed.

Such a timely virtual container incorporates at least three of the process primitives: timing, boundaries, and containers. When designing one we consider the following balancing factors:

• How unexpectedly will the container appear? Will participants be alerted beforehand or confront it by surprise one day? A surprise may frustrate users who want to get on with some other task, but it may also get more spontaneous results.
• How directly will the container appear in their path? Can it be avoided? Is entering it the default or is some action required? Something vital to the group is most effective if it’s the default, but forcing participation may create too much resistance, so some flexibility is useful.
• How many ways of accessing it are there? Is it available through a new menu choice, does it appear in their path, or both? Appearing in the path increases participation, while a menu choice is less obtrusive.
• How quickly and completely will the container disappear? Will it be possible to access it afterwards for archival or other purposes? It can get in the way of the flow of group life later if left in the path. Yet making it disappear too quickly may result in complaints.

By allowing participants to view and respond to each other’s evaluations, we got the intriguing result that every criticism received just as many agreements as disagreements. Opinion was equally split on all issues. For example, some participants complained about lurkers while others said they appreciated being able to lurk or did not mind others lurking. Some participants said would have liked a larger group while others said they would have preferred a smaller, more focused group. The process helped participants realize that the particular balances we had struck in the design were closer to the lively middle then they had realized. The resistance to the boundaries we had drawn was roughly equal.

Shortly after that we created another virtual container using the circle arrangement for a transition ceremony. Several participants were leaving the group and several new ones were joining. During the evaluation someone suggested a ceremony to mark the transition. The one-round ceremony asked those leaving to complete with the group, those joining to introduce themselves, and those staying to share images of their current creative edges — the questions that were most alive for them. Like the evaluation, the transition ceremony appeared on the path into the exchange until a participant responded, and like the evaluation, it was also available through a choice on the menu inside the exchange.

Both the evaluation and transition containers disappeared after a few weeks. We removed them from the path and the menu. It’s interesting to note that no one asked how to get back to the evaluation even though it was a lively discussion. On the other hand, several people asked how to get back to the transition ceremony after it disappeared.

SCRIPTS AS AGENTS FOR GROUP FACILITATORS

In addition to using scripts for such timely virtual containers, we also use them to manage the sequence of events and to maintain boundaries and rhythms. Easily modified as needed, scripts serve as rule-driven, artificial intelligence agents acting on behalf of facilitators, performing functions with accuracy and speed while freeing facilitators to handle subtler tasks requiring human judgement.

Others have used agents as part of computer-mediated communications systems. Hiltz, Johnson, and Turoff (1982) used an “iterative group preference aggregator” to gather opinions from group members, compute a group mean, and then ask those with the most divergent opinions to share their points of view with the group. Object Lens (Lai & Malone, 1988) provides agents for filtering the flow of communications and managing other routine functions.

• checks to see if they have viewed all the guidelines
• welcomes them to the exchange and displays the quote of the week
• checks for the existence of timely special-purpose containers like those for evaluation and the transition ceremony, whether participants have responded, and if not, puts these containers in their path
• involves them in a brief exchange of asynchronous one-line messages expressing how they are feeling at the moment (since most of the activity in the exchange is in self-selected groups, this gives everyone in the exchange at least this brief contact with everyone else)
• offers them a display reviewing new topics and responses to topics in which they are participating
• displays the Creative Exchange menu which includes choices for reviewing the guidelines, participating in any special-purpose evaluation or ceremony, raising a new topic, participating in an existing topic, leaving the exchange, and other actions
• displays a friendly parting message on their way out of the exchange

We also use an agent to manage circle rhythms. It encourages participants to keep up with the group by reminding them to view new items or respond to current rounds. While its behavior varies according to certain exchange settings as detailed in the following sections, it normally encourages without forcing completion of tasks in order to provide flexible rhythms. By its design it manages turn taking, inviting response while making sure that each person responds only once.

VIRTUAL CIRCLE ^(tm)

At the heart of each of our workshops is a Virtual Circle. It is based on a group process found in many native cultures. For millennia, people have gathered in circles around a fire and passed a “talking stick” or other object symbolizing the sacred. As the object passes from hand to hand, each person takes a moment to share his or her truth while others listen with respect. There may be several rounds in a sitting. Each person responds once and only once to each round.

This egalitarian pattern and the time-honored spiritual meaning of the circle drew us to adapt the circle process to hyperspace. We wanted to explore ways of creating equal participation in groups by encouraging lurkers to express themselves. We also wanted to create a reflective atmosphere, to persuade people to take time to discover what really matters to them and then express it in the respectful presence of others.

Our Virtual Circles are supported by a complex of software tools that includes the following components:

• 12 exchange settings which allow dynamic reconfiguration of the circle process
• 3 group markers which define the circle rhythm boundaries
• 3 circle arrangements which represent the most frequently used configurations
• a circle agent action written in C which facilitates the process
• a circle graphic that displays member names in a circle
• boundary agents tailored to the needs of each circle group that manage transitions in and out of the circle container
• background agents that move group markers according to schedule

SPATIAL CIRCLE BOUNDARIES

The membership of a circle is clearly defined. Except for special cases such as dropouts there are few, if any, lurkers. Circle dynamics work best with at least seven and no more than sixteen participants. Groups of less than seven do not work well. Like a chain-reaction with subcritical mass, they rarely get enough interaction going. Groups of more than sixteen overwhelm participants with sixteen or more times as much text as they enter.

Participation is organized into weekly rounds. Everyone responds once and only once to each round. Each week the round consists of one or more evocative questions or self-discovery processes. There are no right or wrong answers to these questions. Their purpose is twofold: (1) to focus the group on a shared issue, and (2) to invite participants to consider what is true for them. Participants often connect early in the week to read the question and live with it for several days before responding in the circle. Figure 8 diagrams the pattern of a question and responses for a single round.

CIRCLE RHYTHMS

Most computer-supported meetings have skewed participation patterns. They also have irregular, arrhythmic distributions when plotted over time. The typical pattern has bursts of activity separated by periods of relative inactivity. Figure 11, a plot of responses over time in creative talk, shows this pattern.

The gate is one of several markers we use to indicate temporal boundaries of current group activity, its “rolling present” or current response window. Our circle process patterns successive windows into a rhythm. In our early circles we assumed that the response window was the current round. While this is approximately true, it’s not quite that simple. In order to shift the group into a rhythm without forcing it, thus creating resistance, we needed several other tools in addition to the gate.

Figure 13 shows our current set of circle management tools in the context of a group “heartbeat” of rhythmic rounds with responses clustering before each gate. It also shows the response window, which may or may not correspond to the current round, because responses to previous rounds are permitted in some situations. In addition to the gate, two other markers are shown, the default marker and the response marker.

The response marker separates previous rounds from current rounds. Current rounds are those which are currently active. While there is usually only one current round, sometimes there may be several, ranging from the response marker up to the gate.

In some arrangements current rounds are defined differently for each member; in others they are the same for all members. In the first case, a member’s response marker indicates the first current round for that member. In such arrangements members who fall behind the group may have several current rounds waiting for their response. In contrast, other arrangements use a group response marker so that all members share the same current rounds. In those, no one falls behind the group; missed rounds are skipped.

THE CIRCLE AGENT

The circle process is facilitated by an agent written in C. The circle action invokes this agent from within a script. A separate script agent periodically invoked by the operating system handles the movement of the gate and other group markers. That script contains a schedule of rounds that can be easily modified by the group facilitator if necessary.

The circle agent uses the exchange settings listed in Table 4, which specify details of the particular arrangement for each circle.

The circle agent guides the user through any previous rounds, inviting him to complete them before participating in current rounds, following the procedure shown in the pseudo-code of Table 5. Then it guides him through as many current rounds as he is interested in, using the procedure in Table 6. Table 7 lists the rules used to determine which rounds are previous or current.

ITERATIVE DEVELOPMENT OF CIRCLE MANAGEMENT TOOLS

During the past two years we have learned through experimentation and participant feedback that effective circles are not as simple as we originally thought. Our action research has taught us much about the appropriate balance between context and form in our circles. As a result we have added settings and markers to give us the functional capacity we need. We now have three circle arrangements which serve different purposes.

Our first design required everyone to respond before the system automatically moved the gate to the next round. It also required everyone to respond before viewing others’ responses, embodying the assumption that participants should share what was true for them rather than being influenced by what others said.

Our first circle involved a dozen of our friends and colleagues who volunteered to try group circle rituals via computer. Several reported feeling anxious about having to respond before being able to view what others had said. To give them more choice, we implemented the rbv (respond before viewing) setting and turned if off so that the system gave them the option of viewing others’ responses before responding themselves (see Table 5).

Several participants went on business trips and got behind, holding up the group while the computer waited for them. In response we implemented the gtr (gate moved when all respond) setting, turned it off, and began moving the gate according to a weekly schedule.

Participants took turns writing the questions for each round. One week, the writers invited the group to create a story together through our responses. The final lesson of our first circle came when two members composed responses at the same time and the story branched in two directions! We then implemented the scn (simultaneous composition not allowed setting) and turned it off since no further stories were planned.

During the Right Livelihood workshop several participants fell behind the group and expressed interest in participating in the closing ceremony without having to respond to the rounds they had missed. We implemented the cnr (continue without responding) and cnv (continue without viewing others’ responses) settings and turned them on, allowing participants to skip over intermediate rounds, go to the closing round, and return later to the skipped rounds.

In another circle a few months later several participants decided to view responses using actions rather than through the circle process and then got confused when responses were displayed again in the circle. To avoid confusion and to prevent members from lurking without participating, we implemented the vop (view outside group process) and rop (respond outside group process) settings and turned them off.

In response to requests from several members of our community, we set up a healing circle where members enter rounds requesting healing for friends or loved ones. In this circle, members sit asynchronously and pray, meditate, or send loving energy to those for whom healing is requested and then enter a few supportive words in response to each round. The group process in such a circle is different than in a workshop. Because members enter rounds (requests for healing) at any time, several of which may be current and active at once, we needed a gate that would move automatically as each round was added. We implemented the gtc (gate moved when round created) setting and turned it on for the healing circle. This pattern of settings became our second circle arrangement, the open circle.

In most asynchronous conferencing systems participants who fall behind are given the opportunity to read everything they missed when they return. Our original circle design reflected this approach. The circle agent would guide the returning user through viewing and responding to all rounds she had missed. However, after getting complaints from overloaded and guilt-ridden participants, we realized that this apparent benefit of the medium was a potential problem.

In face-to-face meetings, there is no easy way to review and participate belatedly if you miss a session. For busy people this apparent disadvantage may actually be a benefit in some situations. They don’t have to review a full transcript to catch up; they simply find out the important points, attend the next meeting, and go from there. Perhaps a rhythmic circle process would benefit from a fuzzy version of the same thing. Perhaps a window during which the round is open but after which the round is moved out of the way would encourage a livelier group.

THREE VARIATIONS ON THE CIRCLE ARRANGEMENT

As a result of evolving our groupware in response to various circumstances, we now have three variations on the circle arrangement:

• A ceremony circle for occasional use. Members respond to the current round or rounds only.
• An open circle for activities such as our Healing Circle. Members create rounds whenever they want and respond to each and every round before moving on to the next.
• A focused circle for our workshops. This is the only circle arrangement with regular, rhythmic rounds.

In a technologically advanced society where production of sufficient goods and services can be handled with ease, employment exists primarily for self-development, and is only secondarily concerned with the production of goods and services....In the “learning society” the occupational focus of most people is learning and developing in the broadest sense. (Harman, 1988, p. 146-147)

The new purpose must be purpose itself. In a “quality of life” reality, the heart of education must be the art of addressing questions of value: learning to identify and ask the important questions, both personal and cultural, and developing appropriate skills to affect one’s world. The new education must be purpose-centered. (Johnston, 1988, p. 6)

What is effective groupware for these times? We believe the most effective groupware helps a group be as creatively whole as possible, to embody more life. It supports diversity of views, synergy, responsibility, trust, and mutual respect. This is achieved by educating (drawing out) and encouraging what is most vital for each of its members. Such purpose-centered education evokes in the individual what is most meaningful and encourages her to live it.

In order to respond to the challenges of our times in effective and sustainable ways, we need to be as creative as possible. This means that each of us brings all of our parts, our primary and secondary processes, to whatever we are involved in together. It is essential to the creative whole that everyone is fully there. We call this whole-person social reality (Johnson-Lenz, 1990). To unleash this creative potency we need to (1) learn new ways of educating ourselves that evoke it, and (2) incorporate those ways into the foundation of our culture, including our groupware.

This fundamental shift in purpose is a cultural sea change already recognized in many spheres:

• Donald Michael (1980) describes the new competence for organizations as learning systems.
• Douglas Engelbart (1988) proposes that corporations institutionalize an evolutionary learning process into their structure.
• Willis Harman (1988) writes that our survival depends on redefining the purpose of work from production to self-development.
• Charles Johnston (1988) articulates a new purpose-centered purpose to education itself.
• Ruben Nelson (1989) describes the necessity of learning how to learn and becoming a learning society.
• Robert Theobald (1987) writes of learning to navigate the rapids of change and becoming social entrepreneurs.
• Lester Milbrath (1989) describes a strategy for envisioning a sustainable society by learning our way out.

THE POWER OF PERSUASION

Not the autocracy of a single stubborn melody on the one hand. Not the anarchy of unchecked noise on the other. No, a delicate balance between the two; an enlightened freedom. (J.S. Bach)

Form is freedom’s right hand. (Johnston, 1990, p. 219)

How do we evoke self-development which is at the foundation of such creative group work? How do we design and apply groupware to maximize the aliveness and capacitance of those using it? We believe the answer lies in bridging the major polarities which frame major unresolved issues in the field of computer-supported cooperative work and groupware, creative tensions at its core. We have already identified four major polarities: human-computer, context-form, sustainability-productivity, and autonomy-commitment, and we propose creative capacity and aliveness as referents for guiding our way between these poles in the design and application of groupware (Appendix D). We conclude by adding persuasion as a tool for bridging between freedom and control.

It’s easy to design systems which leave individuals free to self-organize, often at the expense of much significant happening. It’s easy to design systems which determine or control everything at the expense of individual creativity. We are learning that there is a dynamic balance between these two, a vital middle path that is partly embodied in the software and partly in the action of managers and facilitators. Instead of the poles of total freedom or complete control, we are discovering the power of persuasion.

Robert Greenleaf, who coined the phrase “servant leadership” to suggest that the best leaders serve those they lead, describes three kinds of power (Greenleaf, 1978):

• coercive power: using either overt or covert sanctions to impose your will on others
• manipulative power: guiding others by plausible rationalizations into actions they do not fully understand
• persuasive power: guiding others without coercion or manipulation in ways that allow them to choose actions that make sense to them

Sometimes coercion is useful. For example, in our learning community we insist that new members complete all orientation tasks before doing anything else. If they don’t complete them, they can’t participate. Sometimes manipulation is useful. For example, we occasionally invite users into processes they don’t fully understand or put surprise containers in their path. However, in our experience, persuasion works better in many cases. Our groups are livelier and more creative when we are guiding and supporting them in ways that make sense, they have agreed to follow, and that allow them to choose to act or not.

Such persuasion is not merely providing context, it is actively serving the group by leading and guiding, by providing forms and inviting life to flow through them. Its tools are choices, reminders, options, invitations. The more those forms truly reflect the purposes of that life, the more effectively they will evoke its emergence as a primary process.

THE VITAL IMPORTANCE OF FACILITATION

Machines can do almost anything any time. “Living” change has its seasons. Effectively facilitating it requires a keen sensitivity to timing — when to push, when to wait, when to assert, when to question. (Johnston, 1990, p. 213)

Active human facilitation is a key to effective groupware, particularly during the early stages of group evolution. Of course software agents can play an important role, but human judgement is necessary for effectively managing the subtle dynamics of group energy, particularly when the purpose is to evoke the fullest possible capacitance of groups. Facilitating includes knowing when and how to use rhythms, boundaries, and containers.

Rather than controlling the group, the facilitator’s role is to persuade its members into making choices that embody commitment. As the group matures in its capacity to self-form, its members learn to take responsibility for themselves and the group as a whole. Linda Ackerman (1984) describes a flow state manager’s role as removing blocks in the way of group life. We extend that to include providing appropriate rhythms, boundaries, and containers to support the flow, gently persuading it toward its purpose.

Facilitating on-line groups is labor-intensive. It takes time and energy to focus discussions, ask provocative questions, provide supportive materials, and intervene in difficult situations. Our approach also involves tailoring rhythms, boundaries, and containers through scripts, menus, settings, and markers. Most computer-mediated communications systems underestimate the vital importance of human roles in supporting effective group work. Rather than building such roles into the system’s infrastructure, they discount them by using volunteer facilitators or applying software solutions to what are essentially human problems.

Two hidden assumptions hold this pattern in place. First, our culture tends to discount the vital importance of caring roles such as parenting, teaching, and facilitating. We expect them to be done, but don’t want to pay to have them done well. Second, our culture still believes in the technological fix. We expect computers to do far more than they are capable of in spite of evidence to the contrary. Most expert systems depend on labor-intensive knowledge engineering. Even the best information retrieval systems depend on human judgement and discrimination. It is the same with groupware.

Effective groupware, particularly as we use it, depends on labor-intensive facilitation. While software tools help, they change the nature of the labor rather than eliminate it. They make possible what is impossible without them, but they require experience and time to use effectively.

PRINCIPLES OF PURPOSE-CENTERED GROUPWARE

Expanding on our original definition of groupware by incorporating what we have learned in recent years, we offer the following guiding principles for the design, tailoring, and use of groupware that supports a purpose-centered creative process. No doubt these will evolve and mature in the years ahead as we continue our journey as students of groupware.

1. Creative Whole: Groupware is a creative whole including human and computer components in creative relationship. It is the software of a living system, integrating biological and machine components.
2. Tailorable: Ideally the computer software component is tailorable by facilitators and users to support the group’s needs and purpose.
3. Purpose-Centered: Groupware is most effective if it supports the group’s purpose — a complex function of the purposes of its members and the organization in which it is embedded. If the groupware supports that purpose, group life will flow.
4. A Creative Process: Groupware is a creative, dynamic process that, like any living system, changes and grows. New forms are always emerging from the context of existing forms which they replace.
5. Evolving: Groupware evolves through stages. Guiding form is provided by facilitators during the early stages until the group learns how to be self-forming.
6. Bridging: Groupware is a creative bridging, a dynamic balance between at least five pairs of polar opposites: human and computer, context and form, autonomy and commitment, freedom and control, and sustainability and productivity. The appropriate balance in any situation depends on the purpose and stage of development of the group.
7. Inclusive Referents: The most useful referents for evaluating groupware are measures which reflect the living nature of the system and which inclusively bridge the polarities. Capacitance and aliveness are two possibilities.
8. Whole-Person: The capacitance of a group is greatest — the group is most alive — when the capacitance of its members is greatest, when each member is creatively present and participating with all her capacity. Groupware which persuades rather than controls will evoke greater capacitance and aliveness.
9. Post-Mechanistic Primitives: White space, timing, rhythms, boundaries, containers, and procedures form an array of groupware primitives that bridge the context-form polarity in discrete, functionally distinct steps, each of which serves an essential role in the creative process.

In our work we are learning to follow the vital middle path in between groupware as mechanism and groupware as context. Each of these provides part of the answer to our question of what is effective groupware, but neither is enough by itself.

Most human factors work focuses on the user interface, but little is known about asynchronous group dynamics. Human interaction is richer and more complex than anything reducible to explicit transactions and structured knowledge. What are the fundamental organizing patterns of human interaction via computer? What is the deep structure of life in hyperspace?

While part of our work focuses on supportive software, part of it also focuses on the human system without which the whole is incomplete. Expecting the machine to create a supportive environment is impractical. Sensitive human judgement and interaction are essential. The facilitator/manager is just as critical as the software. On the right hand, her task is to choose appropriate forms to guide group energy, particularly at the early stages of group development. On the left hand, her task is to persuade, educate, draw out, and evoke the inherent self-organizing capacity of the group.

THE CREATIVE IMPERATIVE

We are being challenged not just to the creative uncertainty inherent to any next step in growth, but to a whole new relationship to uncertainty....The critical new questions are demanding a bigger view of things. They are demanding that we take responsibility in a reality without external absolutes, to make our passage into our mature adulthood as a species. It’s a big step. (Johnston, 1988, p. 8)

Never before in history have we been confronted as a species with the awesome responsibility of consciously creating the social systems within which we live and work. In the past these were provided by the culture and passed on generation after generation.

Secondary processes are like those operating in the background of a multi-tasking computer. They are whatever is happening but not acknowledged. They are implicit. They are the thoughts, feelings, and images that challenge the current primary process. They operate in the background, beyond the “edge” of awareness until they emerge and engage with the primary process. Both individuals and groups also have secondary processes. Individual secondary processes reveal themselves in body language, dreams, and our psychological projections. Group secondary processes reveal themselves as any behavior that seems inappropriate to the task or that challenges what is happening. An individual may serve the group by embodying as her primary process something which challenges and is secondary to the group primary focus.

The differentiation phase during which form emerges from its creative context is divided into five stages. While a complete discussion of this model is beyond the scope of this paper, we include brief examples of how each stage applies in the creative cycle of a project, a lifetime, and civilization.

• In the pre-axis stage there is no form, only undifferentiated context. It is the creative incubation stage of a project. In a life, it is gestation before birth. It is before civilization emerges.
• In the early-axis stage, form emerges from its context, but the context is still dominant. It is the inspiration stage of a project. In a life, it is childhood. The child is dependent on the parent. It is the first flowerings of civilization in Egypt, Greece, and Meso-America.
• In the middle-axis stage, the form and its context are of equal importance and the creative tension between them the greatest. It is the perspiration stage of a project. In a life, it is adolescence. The child struggles to be independent of the parent. It is the time from Rome through the Middle Ages.
• In the late-axis stage, the context begins to fall away and the form begins to dominate. It is the polishing and finishing stage of a creative project. In a life, it is adulthood. In civilization, it is the Age of Reason and the Industrial Revolution.
• In the transition stage, there is only form. The context has been completely forgotten, but the transition to the second phase has already begun. It is the stage of a project when one lets one’s creation out into the world. It is the mid-life transition into maturity. In civilization, it is the crossroads at which we now stand.

At our current place in cultural evolution, we are also beginning to remember aspects of the wisdom of past cultures that we have forgotten but which we now need for the next step in our evolution. To embrace the mechanistic world view necessary for the development of industrial culture, we “forgot” our essential connection with nature, each other, and the Mystery of life. Such forgetting was necessary to forge our scientific and industrial capacity, but now this blind spot is the source of many of our environmental and social problems (Walsh, 1984).

• It inclusively values each stage of the creative cycle. Each is an essential part of the process. Rather than discounting the mechanistic world view by focusing on the negative consequences of widespread industrialization, it frames it as a necessary stage in cultural evolution and a vital source of power. Similarly, groupware as mechanism is a vital part of the creative whole.
• By encompassing all stages, it reminds us that none is sufficient. All are necessary. While groupware as mechanism is useful, it is not enough by itself, and neither is groupware as context.
• It aids our understanding of the dynamics between form and context during the differentiation phase. Groupware forms do not emerge full blown from nothing. As with any creative process, they emerge gradually through distinct phases with unique dynamics.
• It reminds us that after form has emerged, the process is only halfway through. Groupware designers tend to think implementation completes the process. Often somewhere in the middle of the process of giving form to our groupware, we forget the living context from which it came, in order to focus on the endless details of implementation. But then the creative process continues. After the groupware is released, feedback from users persuades us to remember the living system we are supporting and to embody its evolving purpose in evolving form.

In this paper we identify five major polarities applied to groupware:

• Human-Computer: Groupware includes both machine and human components. Each contributes something unique to the creative whole.
• Context-Form: Effective groupware provides context for self-organization and requires procedural forms.
• Autonomy-Commitment: Individual autonomy is essential for group creativity and so is commitment.
• Sustainability-Productivity: Effective groupware supports both the productive and sustaining aspects of group work.
• Freedom-Control: Effective groupware both supports freedom and uses control.

Bridging is inherently paradoxical. It seeks to connect the truths in two polar opposite and apparently contradictory ideas. But it is more than merely connecting the poles — it brings them together in living, dynamic relationship. Sometimes one is dominant, sometimes the other. The poles frame the arena of interaction, or as Johnston says, “Paradox frames the door to life.” The greatest creative capacity or aliveness of the system is at a constantly changing point somewhere in between (Appendix D).

• Right-hand error: forgetting the truth of the left hand; groupware as mechanism; enforced commitments no matter what; all interactions are explicit
• Left-hand error: forgetting the truth of the right hand; groupware as context; individual autonomy no matter what; no guiding forms
• Compromise error: forgetting the dynamic nature of the creative relationship; settling on a fixed point of truth somewhere in between right and left; specific groupware as universal solution

The concept is worth considering that the organizing power of life, manifest in mind as well as body — for the two are hardly separable — is the truly creative element. Creativity thus becomes the attribute of life. (Sinnott, 1962)

What is effective groupware for these times? As students of groupware, we don’t expect a quick or easy answer. Rather, we seek to ask it well enough to frame our long-range research agenda. We expect to live with the question for decades to come.

Central to this agenda is measuring our progress. How do we know if we are on course? How do we know if our groupware is effective? How do we evaluate our experiments? With this question, our inquiry shifts from product to process. Rather than looking for the ultimate groupware, we begin looking for a tool to measure the effectiveness of any groupware. Such a referent is not only an important tool for research, but more importantly it is a tool for guiding the dynamic, creative process of evolving, tailored groupware. The referent we choose influences the direction of its evolution.

Productivity is the current referent of choice for groupware. Everyone wants to make computer-supported groups more productive. There is good reason for this. Productive groups get more done. And yet, while productivity clearly measures a vitally important aspect of group life, it is not enough by itself. As a measure of group effectiveness, it has several shortcomings. As a measure of quantity, it is insensitive to the quality of what is produced. By itself, it is also insensitive to the problems of overproduction.

An emerging alternative to measuring the productivity of systems is to measure their sustainability. However, like productivity, it is not enough by itself either. Without openness to change it can lead to stagnation. Enforcing sustainable resource use patterns without concern for humane values can lead to unnecessary stress, suffering, and oppression. Without a willingness to engage with the challenging questions it can collapse into a romantic ideal.

Culturally, we need more inclusive referents which balance productivity with sustainability. Neither is enough by itself. Recently a bridging term has come into vogue that would have been considered an oxymoron only a few years ago: sustainable development. Its paradox points to a middle path which is both sustainable and evolving.

Both productivity and sustainability measure a part of what makes groupware effective. While sustainable development may be useful for environmental issues, something more appropriate to groupware is needed here. What one or two words encompass both right-handed groupware as mechanism and left-handed groupware as context? How can we capture in a single term the essence of what makes groups effective? Johnston (1990) suggests two: capacitance and aliveness. He defines capacitance as the capacity to create. It’s a lot like productivity, but it measures something more fundamental than the output of the system. Aliveness is a straightforward measure of the vitality of a group.

Measuring groupware with referents like these is not as easy as using form-defined referents like productivity. Sales volume, widgets produced, papers published, decisions made, and other objective measures of productivity are easily observed matters of fact. But they are insensitive to the quality of the widgets, papers, and decisions. In contrast, while some intuitive “feel” for a system is needed to know how “alive” it is, such a referent is sensitive to quality. Aliveness is not merely superficial energetic activity. Sometimes silence is more alive than noise, reflection more vital than action. Measuring aliveness takes practice and involves a good measure of subjective judgement, an inevitable consequence of shifting from a mechanistic world view to a creative, living one.

Capacitance is sensitive to the dynamic balance between polarities. For example, maximizing group productivity may work in the short run, but not for the long haul without sustainable patterns which may require lowering short-term output. However, if a group’s purpose is short term, sustainability may be irrelevant. On the other hand, sustainable patterns by themselves can degenerate into holding patterns which are not particularly creative. The appropriate balance between production and maintenance varies from group to group. A group’s capacitance is greatest at a continually changing point somewhere in between, depending on the situation.

• The group gathers within a form provided by a leader or facilitator. Participants follow the rules provided and generally avoid disagreement and conflict. They want acceptance.
• Disagreements and conflicts emerge between participants and between the group and facilitator. This is the most critical stage. Participants may leave. Struggles for power may develop. The group may disband.
• The group learns how to use disagreement and conflict creatively. The capacitance and aliveness of the group is greatest at this stage.

Based on Johnston’s cycle, we expand the three stages and adapt them into our stages of asynchronous group evolution. This model expands on the groupware tailoring process shown in Figure 2 in Part I.

0. No form. No group life. Empty context.

1. Facilitators convene the group for a purpose and provide initial form to coalesce individual energies into a coherent pattern that supports that purpose.

2. Facilitators adjust the form in response to observation and group feedback.

3. Members begin to assert their power by resisting and negotiating the form with facilitators.

4. Members gain sufficient skill to tailor the form to the group’s original purpose with assistance from facilitators.

5. Members gain sufficient skill to change the group’s purpose to reflect the purposes of its members, what is most alive for them. They learn how to use their differences as a source of creative power. They work together to design and tailor forms that support their evolving purpose. The group is a self-forming confluence.

While stage 2 is appropriate for most purposes, including typical business and educational groups, we are entering a time in our cultural evolution when the capacity for groups to be self-creating is increasingly important in all spheres. Network organizations with distributed decision making are more resilient and adaptive to change. Corporations are looking for ways to become learning organizations (Kiechel, 1990) to increase their capacitance.

By definition, a stage 5 confluence embodies the greatest possible capacitance and aliveness of its members, so it is the most creative and potent stage of group evolution, and therefore the most capable of responding to the profound challenges we are facing today. It is the realization of the open space ideal of a self-organizing group. It is a mature stage that few on-line groups have the capacity to achieve or sustain. It is an ideal toward which we reach by growing through stages 1-4, using the “training wheels” of the form provided until we get our balance.

Everything that ever happens in an organization, including the implementation of computer systems, is an example of the “social construction of reality.” Each organization member strives to achieve their own desired goals, which they believe contribute to the overall objectives of the organization. Each individual’s beliefs regarding the connection of means to ends constitute a different version of “reality.” This is the ground on which organizational politics is played out. (Scheer, 1990, p. 99)

All groupware incorporates some elements of group culture: hidden assumptions, guiding myths, values, styles, norms. Of particular interest here are those values having to do with the balance between freedom and control.

Most systems embed the values they espouse in the form or primary process of the groupware — in the transactional rules and who-to-whom networks. However, hidden assumptions are often present in the broader context within which that form arises. To truly support the entire creative process, groupware designers need to be conscious of values in the context as well.

• For our circles (2) to work, participants (1) must commit and intend to participate regularly within the response window, else they will resist. For this to work, our promotional process and guidelines (0) must prepare participants for what is actually going to happen.
• The form (2) of our circles is most effective when all or nearly all members are present. For this, participants (1) must want to express themselves authentically, else they will lurk or be superficial. To prevent this and to encourage members to be present, group norms (0) must allow brief, authentic responses such as “I pass” for those who choose to say nothing.
• For a commitment for action (2) to be fulfilled, the person accepting it (1) must intend to fulfill it and follow through, else breakdown will occur. To assure sincere intent and allow for inevitable changes in one’s ability to fulfill, the groupware (0) must support saying no and negotiating changes.

Bateson, G. (1972). Steps to an Ecology of Mind. Chandler Publishing Company.

Bateson, G. (1979). Mind and Nature: A Necessary Unity. E. P. Dutton.

Bay, D. (1986). Sense of the Whole: Computer Support for Collaboration. Presented at the 16th Annual Information Exchange of the Organization Development Institute, Williams Bay, Wisconsin.

Benston, M. (1990). Participatory Design by Non-Profit Groups. Proceedings of the 1990 Participatory Design Conference. Computer Professionals for Social Responsibility.

Berman, M. (1981). The Reenchantment of the World. Cornell University Press.

Corbin, H. (1976). Mundus Imaginalis or the Imaginal and the Imaginary. Golgonooza Press.

Dossey, L. (1982). Space, Time and Medicine. Shambhala Publications.

Dyson, E. (1989). Groupware. Whole Earth Review, Fall 1989, pp. 105-108.

Engelbart, D. (1962). Augmenting Human Intellect: A Conceptual Framework. Stanford Research Institute.

Engelbart, D. (1988). Bootstrapping and the Handbook Cycle. Presented at CSCW ’88, Conference on Computer-Supported Cooperative Work. Portland, Oregon.

Fischer, G., Girgensohn, A., Lemke, A. & McCall, R. (1990). Conceptual Frameworks and Innovative System Designs for Participatory Design. Proceedings of the 1990 Participatory Design Conference. Computer Professionals for Social Responsibility.

Fuller, R. B. (1969). Operating Manual for Spaceship Earth. Southern Illinois University Press.

Gibb, J. (1978). Trust: A New View of Personal and Organizational Development. Guild of Tutors Press.

Gibson, W. (1984). Neuromancer. Ace Science Fiction Books.

Greenleaf, R. (1978). The Leadership Crisis. Humanitas, November 1978.

Greenleaf, R. (1979). Servant Leadership. Paulist Press.

Harman, W. (1988). Global Mind Change: The Promise of the Last Years of the Twentieth Century. Knowledge Systems.

Hawken, P., Ogilvy, J. & Schwartz, P. (1982). Seven Tomorrows: The Potential Crises That Face Humankind — and the Role of Choice in Determining the Future. Bantam.

Henderson, H. (1981). Politics of the Solar Age: The Alternative to Economics. Doubleday.

Henry, W. E. (1985). In EIES conference C374, “The Social Dynamics of Teleconferencing.” Computerized Conferencing and Communications Center, New Jersey Institute of Technology.

Hiltz, S. R. & Turoff, M. (1978). The Network Nation: Human Communication Via Computer. Addison-Wesley.

Hiltz, S. R., Johnson, K. & Turoff, M. (1982). The Effects of Formal Human Leadership and Computer-Generated Decision Aids on Problem Solving Via Computer: A Controlled Experiment. Research Report #18, Computerized Conferencing and Communications Center, New Jersey Institute of Technology.

Hiltz, S. R. (1984). Online Communities: A Case Study of the Office of the Future. Ablex Publishing.

Hiltz, S. R. (1986). The Virtual Classroom: Using Computer-Mediated Communication for University Teaching. Journal of Communication, Spring 1986, pp. 95-104.

Jacobson, R. (1989). Virtual Places: Strolling the Electronic Neighborhood. Presented at the Second Spirit of Place Symposium, San Francisco, California.

Johansen, R., Vallee, J. & Spangler, K. (1979). Electronic Meetings: Technical Alternatives and Social Choices. Addison-Wesley.

Johansen, R. (1988). Groupware: Computer Support for Business Teams. The Free Press.

Johnson-Lenz, P. & T. (1980). Groupware: The Emerging Art of Orchestrating Collective Intelligence. Presented at the World Future Society’s First Global Conference on the Future, Toronto, Canada.

Johnson-Lenz, P. & T. (1981a). Consider the Groupware: Design and Group Process Impacts on Communication in the Electronic Medium. In Hiltz & Kerr, Studies of Computer-Mediated Communications Systems: A Synthesis of the Findings. Research Report #16, Computerized Conferencing and Communications Center, New Jersey Institute of Technology.

Johnson-Lenz, P. & T. (1981b). The Evolution of a Tailored Communications Structure: The TOPICS System. Research Report #14, Computerized Conferencing and Communications Center, New Jersey Institute of Technology.

Johnson-Lenz, P. & T. (1982). Groupware: The Process and Impacts of Design Choices. In Kerr & Hiltz, Computer-Mediated Communication Systems. Academic Press.

Johnson-Lenz, P. & T. (1988a). Synapse: The Awakening Technology Integrated Conferencing and Knowledge Software. Awakening Technology, Lake Oswego, Oregon.

Johnson-Lenz, P. & T. (1988b). Computer Support for Some Forgotten Aspects of Cooperation. Research Report #1, Awakening Technology, Lake Oswego, Oregon.

Johnson-Lenz, P. & T. (1989a). Connecting with the Heart of Matters in a Virtual Learning Community. Presented at the Electronic Networking Association Conference, “Groupware: The Next Wave,” Allentown, Pennsylvania.

Johnson-Lenz, P. & T. (1989b). Computer Support for Sustainable Culture. Research Report #2, Awakening Technology, Lake Oswego, Oregon.

Johnson-Lenz, P. & T. (1989c). Humanizing Hyperspace. In Context, Fall 1989, pp. 52-57.

Johnson-Lenz, P. & T. (1990). Islands of Safety for Unlocking Human Potential. Proceedings of the Third International Guelph Symposium on Unlocking Human Potential via Computer-Mediated Communication. University of Guelph.

Johnston, C. (1986). The Creative Imperative. Celestial Arts.

Johnston, C. (1988). Beyond Knowledge — Toward Wisdom. Noetic Sciences Review, Winter 1988, pp. 4-11.

Johnston, C. (1991). Necessary Wisdom: Meeting the Challenge of a New Cultural Maturity. ICD Press/Celestial Arts.

Kapor, M. (1990). Quoted in K. Wright, The Road to the Global Village. Scientific American, March 1990, pp. 84-94.

Kelly, K. (1989). Virtual Reality: An Interview With Jaron Lanier. Whole Earth Review, Fall 1989, pp. 108-119.

Kiechel, W. (1990). The Organization That Learns. Fortune, March 12, 1990.

Lai, K. & Malone, T. (1988). Object Lens: A “Spreadsheet” for Cooperative Work. Proceedings of the 1988 Conference on Computer-Supported Cooperative Work. Association for Computing Machinery.

Markley, O. W. & Harman, W. (1982). Changing Images of Man. Pergamon Press.

Meadows, D. H., Meadows, D. L., Randers, J. & Behrens, W. W. (1972). The Limits to Growth. Signet.

Michael, D. N., Larson, M. A., Van der Horst, B. & Wilson, I. (1980). The New Competence: The Organization as a Learning System. Technique Report, Values and Lifestyles Program. Stanford Research Institute.

Milbrath, L. W. (1989). Envisioning A Sustainable Society: Learning Our Way Out. SUNY Press.

Mindell, A. (1985). The Dreambody in Relationships. Routledge & Kegan Paul.

Naisbitt, J. (1982). Megatrends: Ten New Directions Transforming Our Lives. Warner Books.

Naisbitt, J. & Aburdene, P. (1990). Megatrends 2000: Ten New Directions for the 1990s. Morrow.

Nelson, R. F. W. (1989). Learning to Sustain Learning: The Key to Our Future, An Exploration of Personal, Organizational and Societal Learning. Working Paper No. 8, The Post-Industrial Future Project, Square One Management, Ltd., Canmore, Canada.

Nilles, J. M., Carlson, F. R., Gray, P. & Hanneman, G. J. (1976). The Telecommunications-Transportation Tradeoff: Options for Tomorrow. Wiley-Interscience.

Owen, H. (1987). Spirit: Transformation & Development in Organizations. Abbott Publishing.

Peck, M. S. (1987). The Different Drum. Touchstone, Simon & Schuster.

Peters, T. (1987). Thriving on Chaos: Handbook for a Management Revolution. Knopf.

Reder, S. & Schwab, R. G. (1988). The Communicative Economy of the Workgroup: Multi-Channel Genres of Communication. Proceedings of the 1988 Conference on Computer-Supported Cooperative Work. Association for Computing Machinery.

Rheingold, H. (1987). Virtual Communities. Whole Earth Review, Winter 1987, pp. 78-80.

Richman, L. S. (1987). Software Catches the Team Spirit. Fortune, June 8, 1987, pp. 125-136.

Ring, K. (1989). Near-Death and UFO Encounters as Shamanic Initiations: Some Conceptual and Evolutionary Implications. ReVISION, Winter 1989, pp. 14-22.

Roszak, T. (1979). Person/Planet: The Creative Disintegration of Industrial Society. Anchor Books.

Rumi, J. (1984). Open Secret: Versions of Rumi, trans. J. Moyne & C. Barks. Threshold Books.

Scheer, P. (1990). Organizational Politics and Participatory Design, or Why Projects Fail. Proceedings of the 1990 Participatory Design Conference. Computer Professionals for Social Responsibility.

Schindler, C. & Lapid, G. (1989). The Great Turning: Personal Peace, Global Victory. Bear & Company.

Schrage, M. (1990). In Information Technology, the Key Is Good Relationships. The Wall Street Journal, March 19, 1990.

Schumacher, E. F. (1973). Small is Beautiful: Economics as if People Mattered. Blond & Briggs Ltd.

Sculley, J. (1987). Odyssey: Pepsi to Apple...A Journey of Adventure, Ideas, and the Future. Harper & Row.

Sinnott, E. W. (1962). Matter, Mind, and Man: The Biology of Human Nature. Basic Books.

Theobald, R. (1987). The Rapids of Change: Social Entrepreneurship in Turbulent Times. Knowledge Systems.

Theobald, R. (1990). Unpublished outline of The Compassionate Era. Forthcoming.

Trigg, R. H., Suchman, L. A. & Halasz, F. G. (1986). Supporting Collaboration in NoteCards. Proceedings of the 1986 Conference on Computer-Supported Cooperative Work. Association for Computing Machinery.

Toffler, A. (1980). The Third Wave. Morrow.

Turoff, M. (1971). Delphi and its Potential Impact on Information Systems. 1971 AFIPS Conference Proceedings.

Turoff, M. (1982). Management Issues in Human Communications via Computer. In R. Landau, J. H. Bair, & J. H. Siegman, eds., The Management and Evolution of Electronic Office Systems, p. 255. Ablex Publishing.

Turoff, M., Hiltz, S. R., Foster, J. F. & Ng, K. (1989). Computer Mediated Communications and Tailorability. Proceedings of the Twenty-Second Annual Hawaii Conference on Systems Science. IEEE Computer Society.

Wallace, R. (1985). Building the Bridge to Peace: Guidelines for Conflict Resolution. In Context, Spring 1985, pp. 27-32.

Walser, R. (1990). Elements of a Cyberspace Playhouse. Forthcoming in Proceedings of the 1990 National Computer Graphics Association Conference.

Walsh, R. (1984). Staying Alive: The Psychology of Human Survival. New Science Library.

Wheeler, H. (1987). The Virtual Library: The Electronic Library Developing Within the Traditional Library. Published only on diskette by the Martha Boaz Endowment, University of Southern California Library.

Winograd, T. (1986). A Language/Action Perspective on the Design of Cooperative Work. Proceedings of the 1986 Conference on Computer-Supported Cooperative Work. Association for Computing Machinery.

Wright, K. (1990). The Road to the Global Village. Scientific American, March 1990, pp. 84-94.

Zinn, K. L., Parnes, R. & Hench H. (1976). Computer-Based Educational Communication at the University of Michigan. Proceedings of the 31st ACM National Conference.

Zipf, G. (1949). Human Behavior and the Principle of Least Effort. Addison-Wesley.

Zuboff, S. (1988). In the Age of the Smart Machine: The Future of Work and Power. Basic Books.