The problem of increasing societal variety is described. A suggestion is made that social networks might serve as decentralized regulators of
this variety. Examples of social networks serVing in this capacity are given. Methods are outlined for facilitating these networks by sharing information about the
network and its members.
Facilitation at various levels of recursion is discussed.
Computerized conferencing is cited as a means for enhancing communication within geographically dispersed networks. Decentralized computer processing networks are mentioned as the logical hardware counterpart to support these social networks.
A suggestion is made that social networks might serve as decentralized regulators of
this variety. Examples of social networks serVing in this capacity are given.
Methods are outlined for facilitating these networks by sharing information about the
network and its members.
We are living in times of incredible change. Scientific knowledge is doubling every ten years, individuals have increasing personal freedom in lifestyle, our technology brings us new advances and new side-effects, and rapid communications media show us problems and possibilities faster than we can assimilate them.
The increasing variety of problems and options is a blessing and a curse. The variety of new information available makes it more likely that we can find solutions to given problems - if we can find the right piece of information when we need It. The variety of personal options leads to increased freedom, but there is no strong trend toward increased responsibility to go with It. Often the governance and education systems seem to be out of phase with the changes, so their responses are not always appropriate to current situations. How can we cope with this variety ?
A cybernetic principle
One of the fundamental principles of cybernetics, Ashby's Law of Requisite Variety, states that the regulator or governor of a system must match the variety of that system in order to control it (1). Either the variety must be reduced or the regulator expanded until there is a balance.
What Ashby's Law says is that we have a choice. We can reduce the social variety by increasing government surveillance and control, by centralizing our decision-making processes even further, by limiting our personal and collective lives, and by restricting information and research. Many would resist such increased control and limitation, and force would be necessary to maintain it. On the other hand, lNe can increase the variety in our regulatory system by facilitating the free exchange of information, by involving more people in the decisions which affect them and in which they have an interest, by decentralizing institutions, by encouraging localism, and so on. Although more acceptable to most people, this approach must rely on distributed power and governance. and It requires more individual responsibility (instead of dependence on government) for it to work. The political problems and implications of such decentralization are not discussed here.
Sprouts from the grassroots
One promising trend toward increasing the variety of our regulatory systems is grassroots involvement. In many places, people are coming together In loosely organized groups to make sense of and help direct the course of change in their personal and community lives. There Is a resurgence of neighborhood feeling and concern in many parts of the United States. Neighborhood associations are forming to participate in planning, to deliver services to residents, to provide local support, and to participate in local governance. There are also coalitions and alliances of neighborhood associations and other community self-help groups (2]- . Citizen participation and involvement is becoming more prevalent, and is even mandated in some places. Interdisciplinary «invisible colleges of scientists and professionals are forming to share Ideas. There are many public interest and environmental groups which focus on issues they believe important and that government seems to ignore. There are also groups devoted to interpersonal support and personal growth .
The network concept is central to this trend. Many people devoted to alternatives and social change use the term network to describe their group and the relationships and flow of information within it. To them, it means a decentralized network with low centrality, where information passes quite freely among the members and is available to within the network. Furthermore, in this context the term generally includes the idea that power is shared, that decisions are made by all those affected, that economic and physical energy is available to all. In groups with a more collective orientation, there is a notable absence of hierarchical structure, and authority Is often split to assure that the ideas of anyone person do not dominate. Many people involved in social change and innovation proudly call themselves 'networkers'. They are well practiced in the network arts : sharing information and leads to other people, helping bring people together who can mutually benefit, helping people find what they need .
Decentralized social change networks based in the grassroots constitute a promising beginning for a change in our governance system that has the potential for matching the variety of our time. They are especially powerful because they are grounded in people's personal lives and the friand1ilhip networks that make up our social fabric. They can begin to match the variety of problems, needs. resources, and conditions as their memberships and purposes change in response to the changing times. Being fleXibly structured, they can respond more quickly than the more rigid social institutions of today . If necessary, an entirely new network can emerge from the pieces of an old one. These networks can also target their responses to the appropriate places, with the appropriate levels of help. They can bring to bear many diverse talents. Being rooted In the people, they can bring local understanding to local problems which bureaucrats don't always share.
Because of limited communications channels within and among themselves, these networks cannot always respond quickly and easily to problems and issues. Communication is often limited to sharing information through the mail, printed newsletters, and occasional telephone calls, whenever face-to-face meetings are not possible. This is a serious problem in geographically dispersed networks, such aS the loosely 'Organized Northwest Net. It includes perhaps a thousand people who are working on local food production and distribution, alternative and public acceBll media, holistic health, land trusts, communications, and more In various subnetworks in Oregon, Washington, and Idaho. These networks are further hampered by the slowness of the natural word-of-mouth process by which people come into 8 network and find others with common interests. Such limitations make it difficult for these networks to evolve into a meta-network of issue-specific ad hoc groups emerging in response to issues and then fading away as the problems are solved.
If these networks are to develop further in the direction of regulating life on the planet, they must be facilitated. Their capacity to link members and to communicate with other networks must be enhanced. This is the motivation for our work, as well as the work of others interested in the birthing of new planetary regulatory systems. Our own work consists in using the tools of the communications era (computers, telecommunications, mathematical models and methods, etc.) to increase the ability of these networks to perceive problems, to link up into adhocracies for action, and to interconnect with other networks. Facilitating networks involves distributing information about the network to all its members. This information includes facts about members' skills, resources, needs, availability, attitudes, interests, and perceptions. It may also include information about the structure of the network. By sharing as much "access". information as possible within a network, individual members are empowered to form their own links with others, without having to rely on a central leader. By sharing information about members' perceptions, or "mental models", it becomes easier for subgroups (or subnets) to form for discussion or action. The purpose of network facilitation is to increase the number of links among members and to decrease the degree of centrality of the network.
Facilitation through sharing information about people
Many of our projects have been based on building a file of information about people in the network, containing the names, addresses, telephone numbers, and some additional information about concerns and interests. This additional Information may include both keyword descriptors and free-form textual material.
The International Network for Social Network Analysis (INSNA) directory we prepared is a good example of such a file . Even in print form, this information allows INSNA members access to everyone else in the network. The keyword indices provide a way to locate others in the same discipline or geographic area, or those with similar interests. The INSNA directory is now available on a computer at the University of Toronto. The on-line file can also support more complex searches; for example, for people in Canada who are sociologists, and who are interested in support networks and methods for investigating them.
By adding more descriptors for each person, more refined searches become possible, including searches based on c profiles. or sets of characteristics. The development of keyword descriptors for people in a network should be done with the advice and consent of network members. There are serious problems with an open-ended list of keywords. First, if participants make up their own descriptors, duplicate keywords with slight variations often occur. For example, one might use « gardener », while another would say « gardening ». Second, synonyms or closely associated terms often appear as separate keywords, such as "women's studies", " women's movement ", "feminist movement", and so on. An initial keyword list may be developed by a network organizer or facilitator, but network members should be asked if those keywords describe them adequately and what changes should be made. There should also be provisions for adding or modifying descriptors as the network changes.
Another way to bring people together in a network is to share information about members' points of view about given topics. Recent developments in modeling theory (inclUding interpretive Structural Modeling) have produced techniques for structuring the elements and relationships that make up a person's view of a topic into an integrated mental model . Using directed graphs, a person's mental model can be expressed as a network of concepts. Rather than using ISM techniques which produce a single group model, we have chosen to ask each person questions about the elements and relationships he or she perceives and then to "cluster" the responses into patterns (using noway tabulations to find exact pattern matches). Then the most frequent patterns of responses (that is, the most frequent c mental models.) are shared with network members. Not only does this tell members what points of view they and others hold, but it also provides an explicit opportunity for discussing points of difference. We generate the initial list of elements and the possible relationships among them with a small, diverse group of people familiar with the area or issue.
We recently used such techniques at the Oregon Information and Referral Idea Fair and Workshops. Before' the Idea Fair, we generated some initial models of information and referral (I&R) and conducted a pilot test with a diverse group of people involved in community and social serive I&R. Then, at the Fair, following registration, we surveyed the participants, entered their responses into the computer, analyzed the results, and later shared with the participants the most frequent mental models of information and referral. showing not only what they felt about I&R, but why . By using such techniques we are sharing not only a specific interest or attitude, but we are also beginning to make explicit in broad terms the entire constellation of what a person thinks about a given area, 50 that everyone has a contexted picture of what others in the network think about a topic.
Facilitation through sharing Information about networks
Another kind of information that can help people in a network is information about the network structure who knows whom, who has worked with whom, etc. This sort of information is common to most social network analysts, but it is relatively new to social network practitioners. We believe that such data can be used to modify and extend existing social networks. For example, if one joins a network and knows a few people, he or she can use portions of the whole network data to find friends to introduce him or her to other interesting people in the network. Brokering can also be done more formally by people in the network who enjoy match-making. Information about other networks to which one belongs can also be shared in this manner, thus providing linkages among networks through node individuals. In our experience, most changes in social networks are accomplished through existing links; we have been introduced to most of our friends by other friends. Access to whole network data of this type can facilitate the natural process of network growth.
We are also participating in network communication and facilitation on EIES (Electronic Information Exchange System), a computerized conferencing system designed by Murray Turoff. This winter, some members of INSNA will be using EIES to participate in a network of social networkers convened by Unton Freeman. This network will share ideas and work as a geographically decentralized «invisible college " combining several academic disciplines in the discussion of social network analysis  -. Whole network data has been collected on INSNA, and plans are being made to analyze this data to give a better picture of who its members are and how they interrelate. Similar plans are being made for stUdying the network structure of the social networks network on EIES, as well as of other EIES networks. What are the effects on the network structure of making that structure explicit to all the members?
Will it stay the same or change, and how?
Levels of facilitation
Facilitation must take place at several different levels of networks. First, it must happen at the level of each neighborhood and local community. We have designed and used a computerbased community information system to help organize our neighborhood for participation in a comprehensive landuse planning process. We began with a survey of our neighbors. From the survey results we developed the neighborhood's agenda for action and prepared lists of neighbors with similar concerns to serve on task forces and committees. The system was also used to form a telephone tree for communicating and responding to surprise moves from City Hall. The entire effort had a significant impact on political directions in the city . In addition, the system could have been used to bring people together for social purposes, in common interest groups (e.g., gardening club, play reading group, etc.), or to exchange goods and for services. However, the neighborhood association chose to emphasize political and planning issues rather than social organization.
We also helped a project get started in Portland, Oregon, where a neighborhood association is using a microcomputer in someone's basement to facilitate the exchange of skills and resources among nelghbors .
Micro-computer hardware is becoming inexpensive enough ($600 and up) to enable interested neighborhoods and community groups to handle their own information needs without outside assistance. To make this happen, a variety of software packages and people willing to maintain and manage such projects are needed.
Second, facilitation must happen at larger levels of perspective - at the county, state, regional, and eventually national and global levels. At these levels there are several problems: providing communications channels for large numbers of geographically separate people, interllnking and interconnecting more local networks for largescale action, and organizing large-scale complex problems so that the problem components and the relationships among them can be readily understood. At theM larger levels of perspective, many networks and « networking» projects exist. Harry Stevens has been designing and testing techniques for « Involvement through networking » for fifteen years. He is currently developing a Science Resource Network for the Massechusetts Legislature  and planning a legislative exchange experiment among state legislatures via notebooks and computerized conferencing. Last winter we participated in the design and development of a social process and computer system to support city- and state-wide issue dialogues in Washington State -. Issues were formulated and analyzed by citizen groups, who accessed the results through an interactive computer at meetings. These issue dialogues clarified not only who felt which ways about issues, but also why they felt those ways. This can be the basis for organizing into action groups and forming political coalitions. In Hawaii, the Hawaii Health Net links people interested in holistic health . There is a state-wide technical skills bank in North Carolina, and a national skills bank is being developed by Patrick Saccomandi of the Independent Foundation  . On a global scale, Anthony Judge has used the network paradigm to express and interrelate perceived problems, the international organizations concerned with them, the disciplines focusing on them, and the values which make them visible -.
Networks of people also share Information about topics of common interest, goals, purposes, etc. Local networks can often do this in face-to-face meetings, but geographically dispersed networks must circulate textual and graphic material through the mail. This is slow and expensive, and truly interactive. communication is impossible. One solution.to this problem is computerized conferencing, which allows groups to communicate ideas, "meet" and make decisions, without the cost of travel and the inconvenience of bringing people to a central location at a given time . Such conferencing Is asynchronous, since material may be entered into or retrieved from the computer at different times, thus making rapid communication within a network possible at the convenience of each individual. We are aware of several groups of scientists, social scientists, and others interested in social change who are exploring some means of bringing together geographically dispersed people into networks to share ideas, make friendships, and work together. We are helping several of these groups find appropriate state-of-the-art communications systems to support their networking activities. At present, full computerized conferencing systems are not widely available to most networks, but they will be in the future. We feel that the potential for computerized communications systems to link people in dynamic, ever-changing, decentralized networks is virtually unlimited.
In a few more years, people at home will be able to have computer terminals hooked up to their family TV sets for a few hundred dollars. Already, experiments are being conducted with systems in England that will deliver textual information to subscribers' TV screens . In Columbus, Ohio 100,000 homes are now wired for two-way cable TV, which began programming in December. 1977 .
Such communications systems begin to support the variety in society, but they also need to be structured so that the variety is regulated, rather than expanded into chaos.
Next: Decentralized computer networks
Most of the current experiments in social network facilitation using computers have been limited to using a central computer to store the directory for the network, to analyze the structure of a network, and to support computerized conferencing. Even though a single, central computer may be accessed through geographically distributed computer terminals, the current state-of-the-art involves centralizing the data in one place. This centralization has the same shortcoming we mentioned before: it tends to limit variety.
Recently, computer scientists have begun experimenting with 'distributed processing networks'. Such a network is made up of many computers, themselves geographically distributed. The major advantages of such networks are that local processing can be done by a local computer, sensitive data can be kept in a local computer and thus protected, other computers can "help" in a problem when needed, and the activity of the entire network can be dynamically allocated to the current set of problems. Such a decentralized network has no central data base. The data is kept in bits and pieces in the distributed computers. A distributed processing network is the logical hardware counterpart to the social networks discussed above. Loving Grace Cybernetics is currently developing a distributed processing network that will serve as a "community memory" in the San Francisco area, containing information about community needs, services, resources, and so on .
Given the increasing variety in our society, it is necessary to find new mechanisms for coping with it and with rapid change. Either the regulatory systems need to be amplified, or the variety needs to be reduced. Networks of people coming together out of common interest and concern may serve as an adjunct to current regulatory systems to match the exploding variety. Such networks need to be organized and facilitated at various levels of recursion, beginning at the local level. Information about people's interests, mental models, abilities, concerns, values, and so forth needs to be shared within and among networks. Information about the network's structure can also be used to facilitate the development of new relationships within the network.
Geographically dispersed networks of people can be facilitated through new communications technologies, including computerized conferencing. In the future, decentralized computer networks will also play a part. These trends suggest new governance and educational structures that may help us preserve our freedoms, while bringing more individual responsibility to bear on new problems.
References and Access Information
- W. Ross Ashby, Introductlon to Cybernetics (London: Chapman & Hall, 4th imp. 1961), chap. 11.
- David Morris and Karl Hess, Neighborhood Power : The New Localism (Boston: Beacon Press, 1975). David Morris, Institute for Local Self-Reliance, 1717 18th St., NW., Washington, D.C. 20009.
- Milton Kotler, Neighborhood Government: The Local Foundations of Political Life (Indianapolis: Bobbs-Merrill, 1969). Milton Kotler. National Association of Neighborhoods, 1612 20th St.. N.W., Washington, D.C. 20009.
- Brian Livingston, "Neighborhoods and Communities", Cesede: Journal of The Northwest, May, 1977, pp. 4-7. Brian Livingston, Cascadian Regional Library, P.O. Box 1492, Eugene, Oregon 97401.
- National Association of Neighborhoods, 1612 20th St., NW., Washington, D.C. 20009. D.C. 20007.
- ACORN (Arkansas Community Organizers for Reform Now), 523 W. 15th, Little Rock, Arkansas 72202.
- Interpersonal Support Network, 311 California St.. Room 700, San Francisco, California 94104
- Steve Johnson, "Networks" Rainbook: Resources for Appropriate Technology (New York: Schocken Books. 1977). p. 70. Steve Johnson. 20950 S.W. Farmington Rd., Beaverton, Oregon 97005. RAIN-Journal of Appropriate Technology, 2270 N.W. Irving, Portland, Oregon 97210.
- Byron Kennard. "Keep Watering the Grassroots", Environmental Action, July, 1977, pp. 12-14, Byron Kennard, Environmentalists for Full Employment, 1785 Massachusetts Ave. N.W., Washington, O.C, 20036.
- "Membership Directory", Connections: Bulletin of the International Network for Social Network Analysis, I (Summer, 1977), pp. 3-20, INSNA, Barry Wellman, Principal Coordinator, c/o Centre for Urban and Community Studies, University of Toronto, 150 St. George St., Toronto, Ontario M5S 1AI, Canada.
- John N. Warfleld, Structuring Complex Systems (Bettelle MemOrial Institute Monograph No. 4; Columbus. Ohio: Battelle, 1974), John N. Warfleld, Baltelle Columbus Laboratories, 505 King Ave., Columbus. Ohio 43201.
- Peter & Trudy Johnson-Lenz, "Conference Facilitation by Computer-Aided Sharing". Transnational Associations, XXIX (October. 1977), pp. 441-445. Peler & Trudy Johnson-Lenz, 695 Fifth St., Lake Oswego, Oregon 97034.
- Linton C. Freeman, "Computer Conferencing and Productivity In Science", Transnational Associations, XXI X (October, 1977), pp. 434-35. Linton Freeman, Dept. of Social Relations, Lehigh University, Price Hall, Bldg. 40, Bethlehem, Pennsylvania 18015.
- Murray Turch. James Whitescarver, and Starr Roxanne Hiltz, "Assisting Invisible Colleges" by EIES., Transnational Associations, XXIX (October. 1977), pp, 43133, Murray TuroH and James Whitescarver, Computerized Connferencing and Communications Center, New Jersey Institute of Technology, 323 High St., Newark. New Jersey 07102. Starr Roxanne Hiltz, Dept. of Sociology and Anthropology, Upsala College, East Orange, New Jersey 07019.
- Peter & Trudy Johnson-Lenz, "Description of a Neighborhood Information System. (unpublished manuscript), April. 1977.
- APPLE, A Person-to Person Living Exchange, 817 N.W. 23rd Ave., Portland, Oregon 97210. The micro-computer is located at The Life Support Systems Group, Ltd., 2432 N.W. Johnson, Port: land Oregon 97210.
- Chandler Harrison Stevens. "Science Resource Network for Legislators and Citizens", Science and Public Policy, October, 1976, pp. 442-454. Harry Stevens. Participation Systems, Inc., 43 Myrtle Terrace, Winchester, Massachusetts 01890.
- Hawaii Health Net, Moihili Community Center, 2535 S. King 51, Honolulu, Hawaii 96614,
- North Carolina Office of Citizen Participation, 401 N. Wilmington St., Raleigh, North Carolina 27601
- Independent Foundation, 1028 Connecticut Ave., N.W., Suite 618, Washington, D.C. 20036.
- Yearbook of World Problems and Human Potential (Brussels: Union of International Associations and Mankind 2000. 1976). Anthony IN. Judge, Union of International Associations, 1 rue aux Laines, 1000 Brussels, Belgium.
- Anthony J.N. Judge, "The Harmony of Interaction - and the Facilitation of Network Piocesses", International Associations, XXVI (1974), pp. 538-543.
- Murray Turol1 and Starr Roxanne Hiltz. "Meeting Through Your Computer", IEEE Spectrum, May, 1977, pp. 58-64.
- Nicholas Valéry, "Foot in the Door for the Home Computer". New Scientist, April 14, 1977. pp. 63-5.
- "QUBE Fact Sheet" QUBE (Warner Cable Corporation), 930 Kinnear Rd., Columbus, Ohio 43212.
- Loving Grace CybernetiCs. 1807 Delaware St., Berkeley, California 94703
* Originally written for « Connections », the Bulletin of the International Network for Social Network Analysis. Reprinted with Permission.