WPbanner5.jpg (10176 bytes)
No. WP- 02-09
Group Behavior and Learning in Electronic Forums:
A Socio-technical Approach

Rob Kling and Christina Courtright  (11/5/02, v. 5.7)
Center for Social Informatics
Indiana University
Bloomington, IN 47401




To appear in: In Sasha Barab, Rob Kling, and  James Gray (Eds.).Building Online Communities in the Service of Learning.
Cambridge University Press (2003).
 

Introduction
The term “community” is widely and often uncritically used to characterize two kinds of groups that are central to this book. First, there are groups that come together to learn, in classes, workshops, and professional associations. Most professionals would refer to these as classes, workshops, and associations. Some educators like to refer to all of these kinds of groups as “learning communities.” The second kind of group is one that participates in an electronic forum (e-forum), such as an Internet Relay Chatroom, a professional LISTSERV, a distance education course, or an online auction. It has become equally commonplace to refer to such groups as communities —virtual communities. This book examines issues for the developers and participants of electronic forums that could facilitate learning.

We believe that the casual use of the term community to characterize groups that are engaged in learning, or groups that participate in e-forums, is seriously misguided. As we shall see, developing a group into a community is a major accomplishment that requires special processes and practices, and the experience is often both frustrating and satisfying for many of the participants. The extent to which a group develops certain desirable community-like characteristics should be based on empirical observation rather than on assumptions or aspirations.

Since 1999, we have been involved in the development and study of an e-forum designed to support science and math teachers who want to improve their abilities to teach with inquiry approaches. The Inquiry Learning Forum (ILF) is a large project that has directly involved eight faculty members and eight graduate students, with a research advisory board of nine faculty and a participant advisory board of eight teachers. The ILF's e-forums were opened for teachers' access in the Spring of 2000, and by mid-2001, several hundred teachers had registered and visited them.

Groups of this scale and diversity rarely have a consensus about the processes and practices that will support teachers in improving their teaching. We have participated in numerous meetings and informal discussions of the nature of participants’ behavior in the ILF's e-forum (which we call the e-ILF). We have heard the term community used in a wide variety of ways, and have seen important evidence that the “assumption of community” on the e-ILF has undermined our ability to effectively support reflective discussions by the participating teachers.

This chapter examines how group behavior in e-forums could be supported to meet the kinds of aspirations of trust and reciprocity that people who use the term community desire. We will start with a discussion of the structure of the Internet, to address some of the prevailing misconceptions regarding its nature as a medium for human interaction and community development. We will then discuss a variety of social forms, from teams and groups to communities. We will also examine some of the available research concerning how the crucial element of trust is developed in groups, both face-to-face and those that meet or work online. Finally, we will discuss some of the relevant experiences of the ILF, and provide suggestions for structural changes that could enhance its community-oriented features.
 

Contrasting Conceptions of the Internet

Standard Model of the Internet
The growth of the Internet as both a tool and an environment for human interaction has generated a concomitant growth of popular and academic literature seeking to describe, interpret, and explain the nature and forms of such interaction. Underlying much of this literature is a conception of the Internet as a “level playing field” whose architecture allows people to engage in many of the activities that they have traditionally performed off-line, including conversation, work, commerce, hobbies, meetings, worship, reading, and learning, yet without the usual constraints of space and time. This “level” or “standard” model of the Internet is conceptualized most fundamentally as a network of computer networks connected by virtue of the TCP/IP protocol, that allows unlimited sending and receiving of electronic files (Cerf, 2001; Zakon, 2001). What derives from this essentially technological characterization is a portrait of human online activity that mimics the fluid, boundless nature of file transfer: people can communicate with anyone, at any time, in groups or one-on-one; anything can be located, read, or purchased; work can be conducted from anywhere; and so on.  Even the complaint that “there is a lot of unreliable information on the Internet” rests on this standard model.

Socio-technical Model of the Internet
An alternative model conceptualizes how the Internet is structured technically, socially, and socio-technically. The term socio-technical refers to an ensemble, a practice, or an analysis of these that integrates social and technical elements in a way that reveals their interactions and inter-penetration. In conventional approaches, an ensemble could be seen as wholly social, such as an organization, a high school for example. More accurately, an analysis of the high school could be wholly social, and focus on such behaviors as the formations of informal student cliques. Or the ensemble could be seen as wholly technical. The high school could be the subject of a technical analysis of its structures: the likelihood that it would collapse in an earthquake. A socio-technical analysis of behavior in the high school would focus on a mix of social and technical elements. For example, if the classroom chairs are bolted to the floor in some rooms and not in others, one may examine different kinds of class organization and communication in these rooms which different physical capabilities. The structuring of spaces, such as rooms that are assigned to student clubs, may influence the ways that cliques form and how they manage their social boundaries with technologies (i.e., closed doors). High schools are neither completely social entities nor completely technological entities, even though social psychologists and structural engineers may analyze them as if they were. Socio-technical analysts see social behavior and the organization of artifacts (such as buildings or web sites) in a much more integrative manner.

Consider Web sites as varied as newspapers (such as the New York Times online edition), travel sites (such as expedia.com), financial services (such as schwab.com), and e-magazines (such as salon.com). The conventions for these sites vary considerably. They allow their readers few places to post, if any. Some e-zines, such as salon.com, have areas that anyone can read, along with premium areas that require a paid subscription. In contrast, the New York Times online edition requires a potential reader to register in order to read stories that are less than seven days old, and to pay a fee for accessing older stories from the Times' archives. Financial services such as schwab.com provide scant financial information for people who have not established a brokerage account with the firm. Schwab's account holders, however, can enter the site and find extensive reports, charts, and other means for evaluating potential investments. Some of these sites rely upon passwords for protection; others use IP addresses. In short, rather than the level and undifferentiated view of the Internet that emphasizes highways from a user to every site, this socio-technical view emphasizes carefully structured electronic forums where people experience walls, hallways, and doors with electronic locks.

A different class of Web sites is structured to support more open communication among participants, and sometimes high degrees of sociality. Clubs on Yahoo.com and America Online require simple subscriptions. But some of them screen applicants and limit membership to certain classes of people, such as adults, or teenagers, or women. ArXiv.org, a repository of scientific working papers, allows authors to post their own works. Anyone can visit the site and read articles without registering. However, those who wish to post articles must register and demonstrate their legitimacy (by having accounts with .edu or .gov domains). While scientists may post articles on arXiv.org, there are no spaces to discuss articles that have been posted.

Online games support a different kind of sociality. Popular games, such as Ultima Online and Everquest, each have several hundred thousand players paying monthly subscription fees. In order to advance in these games, the players join groups, called guilds. Not only are the spaces structured to help players form guilds, but some game designers also create incentives for players to recruit and assist new guild members by awarding them additional credits and other resources for their efforts. Sanctioned behavior varies among games, from healthy competition to group murder (Ahuna, 2001; Kolbert, 2001).

Thus, sites like these are structured socio-technically, in that they are co-configured not only by the constraints and affordances of the technologies involved, but also —and primarily— by social, economic, and institutional factors. The sites also differ in their rules and norms about:

In our view, e-forums can be designed socio-technically (Kling, 2000). Designers may try technical means (such as IP-address checking or personally-reviewed registration) to limit participation to some kinds of people or groups. An e-forum's charter may describe the kinds of acceptable communications, but without human review, it may be difficult to enforce. The e-forum may enable some specific kinds of activities, while providing no explicit support for others. The e-forum may be designed so that any participants may easily form their own (private) group, or they may require the assistance of an e-forum administrator, or it may be effectively impossible.

These socio-technical design elements may tend to support some kinds of social relationships between participants, but they do not effectively control the array of possible and even likely behavior. In fact, participants in e-forums can —and do— subvert designed structures for their own purposes, just as members of a club or association can push the group to act in a manner not foreseen by its founders.

For example, the BioNet newsgroups (http://www.bio.net/) were established to enable research biologists to discuss specific topics; lists were created in the early 1990s for specialized topics such as automated DNA sequencing, microbial biofilms, and the science and profession of biophysics. By 2001, the unmoderated BioNet newsgroups had become loaded with spam (i.e., advertisements for “easy credit,” “find out anything about anyone”); and some seem to have died as scientific e-forums. A related example: the members of one Usenet newsgroup, alt.tasteless, once decided to “raid” another unmoderated newsgroup, rec.pets.cats, and filled it with posts that the cat lovers found deeply offensive (Quittner, 1994).

Sometimes participants will subvert the e-forum structures in an effort to be constructive. One of the authors (Kling) was invited to give a guest lecture for a two-week period on “social informatics” in a doctoral course that was organized through a set of e-forums. One of his social informatics articles was posted as “a lecture,” and about 20 doctoral students were required to post one or two questions in response. The e-forum was structured with a special area (list) for Questions, and a separate area (list) for Answers. After reading some of the posted questions, it seemed that an online discussion would be more useful than simply posting static answers. Kling thus posted replies, comments, and questions of his own in the Questions area exclusively, and encouraged the doctoral students to respond there as well. He succeeded in developing a discussion that engaged over 40 doctoral students during the two weeks; this would probably not have occurred if questions and answers had been kept to separate realms.

In this socio-technical view, then, the Internet is a “chunky” environment populated by many different kinds of spaces, each structured both socially and technically. Of particular interest to this chapter are the e-forums that support high levels of communication between participants. Groups form around e-forums, rather than simply “on the Internet.” Among such e-forums, no one mode of social interaction predominates. Nor can they be easily aggregated into an umbrella “Internet community.” Instead, just as in face-to-face life, there are forums that resemble fan clubs, for example, while others resemble flea markets, and yet others look like task forces working closely on a project. And some may form communities. Each e-forum has its own norms, purpose, accessibility, and expectations, although many bear family resemblances to a cluster of similar e-forums and their participants. This forum-centered socio-technical approach will frame our examination of the e-ILF, following a more detailed discussion of forms of social organization.
 

Social Organization: From Teams and Groups to Communities
Empirical research in the social sciences has generated a plethora of categories to describe forms of social organization, ranging from teams and task forces to groups and communities. One convenient way to organize these types of social organization is to divide them according to their relationship to paid employment. Work-oriented forms of social organization include, for example, task forces and work teams, trade unions and guilds, professional associations and study groups, classrooms and skills centers. Each has notably different attributes, but what they share in general is their members’ recognition that participation is either compulsory or highly beneficial for work-related advancement.

Non-work groups can be further divided into voluntary associations, which are generally task-oriented even if sociability plays a large role (Sills, 1968), and “hangouts,” or non-task-oriented social formations (Oldenburg, 1999). Among the former can be found, for instance, clubs and non-professional associations, fan clubs, mutual self-help groups, congregations, resource centers, interest groups, and even gangs. Purely social hangouts include, for example, cafés, taverns, bookstores, farmers’ markets, parks, beaches, malls, downtown, water coolers, and hair salons.

Sociological and anthropological studies have carefully examined these formations, and it is not difficult to identify and illustrate “typical” attributes and roles found in each. Communities, however, constitute a much more elusive analytical category, yet the term is used with increasing frequency in both lay and academic literature, particularly with regard to online social formations (cf. Hagel & Armstrong, 1997; Kim, 2000; Preece, 2000; Werry & Mowbray, 2001). We will examine both lay and academic uses of the term in the next section.

Sociological Communities?
Community is a strange and particularly resonant term in North American public life. Like many key concepts in the social sciences —including culture, learning, politics, and power— it has specific and restricted meanings for scholars, and broader connotations when it is used in lay language. As a lay term, community usually connotes a group which shares warm, caring, and reciprocal social relationships among its members. It is invoked routinely by politicians and real-estate developers to set aspirations for places and their residents. As an analytical term, sociologists, anthropologists and others have examined alternative definitions and the complex, multivalent social relationships between people who participate in various communities, including spatially concentrated villages and geographically dispersed professional associations (reviewed in Brint, 2001; Hillery, 1955; Morris, 1996). A major analytical effort by F. Tönnies (1887/1955) that gave a new empirical anchoring to German sociology in the late 19th century distinguished between two broadly-conceived ideal types: village-style Gemeinschaft (community), associated with close ties and shared values, and city-style Gesellschaft (society), characterized by dispersed ties and dissimilar views.

In practice, Tönnies’ categories are essentially mid-level social constructs used to characterize an intermediate level of social organization between individuals (or households) and the totality of a society, whose value is more practical than analytical. In particular, the term community has achieved this practical meaning in everyday life. For example, a call for “community management” of a federal program implies involvement by a potential range of non-governmental actors including non-profits, churches, or coalitions that are local to those who receive services. In this case, “community” is implicitly contrasted with a larger entity, “the state,” in terms of its style of management.
Analytically, however, the term community must be more carefully examined. In a review of empirical studies, sociologist Stephen Brint (2001) has identified six dimensions of community that are well supported by the sociological research literature: 1) dense and demanding social ties; 2) social attachments to and involvements with institutions; 3) ritual occasions; 4) small group size; 5) perceptions of similarity with the physical characteristics, expressive style, way of life or historical experience of others; and 6) common beliefs in an idea, a moral order, an institution, or a group. Brint notes that few groups share all of these characteristics simultaneously. In addition, research has associated each of these six dimensions with other social outcomes and characteristics. For example, dense social ties can be associated with conformity to the dominant morality in the group.

Brint also notes that careful empirical studies of communities do not reinforce their popular image in lay usage. For example, cooperation is not a defining characteristic of communities. In the 1930s and 1940s American sociologists were discovering hidden patterns of privilege, power, and inequality in communities that at first glance seemed cohesive. Sociologists also learned that what seemed like the spontaneous development of group consensus could often be a byproduct of the self-interest of dominant status groups. Nevertheless, not only in lay and professional literatures, but also in scholarly literature, communities are most often represented as a highly desirable form of social relations, characterized by warmth, cooperation, and mutual support. This characterization is particularly prevalent in popular and professional literatures addressing education, business, and online sociability.

When a term is used to depict an ideal or desired state of affairs rather than to analyze an existing reality, it can be considered aspirational. One of the best-known examples is the stirring affirmation in the U.S.'s Declaration of Independence that “all men are created equal,” despite the widespread practice of slavery in the colonies at the time, even among some of the declaration’s authors. Of course, stating aspirations is important (as in the case of Martin Luther King's “I Have a Dream” speech), and sets a sense of direction for a common enterprise. Nevertheless, aspirational definitions are, by intent, empirically inaccurate. Their purpose is not to analyze existing realities, but rather to envision desirable future situations and thereby motivate change from the current conditions. We find that many uses of the term community are, in fact, aspirational rather than empirically grounded. Unfortunately, too many authors do not communicate the scope of the challenge involved in moving from what exists to what is desired. As a result, community-building tends to involve unforeseen levels of hard work, conflict, and renewed efforts at group mobilization to achieve.

Virtual Communities?

Another example of the aspirational use of the term community can be found in the vast literature describing online social interaction. Howard Rheingold (1993), who is frequently cited on the subject, defines virtual communities as

…social aggregations that emerge from the Net when enough people carry on […] public discussions long enough, with sufficient human feeling, to form webs of personal relations in cyberspace. (p. 5)
Our criticism of this definition is that it does not distinguish between the kinds of organizations that are basically like groups, like hangouts, like associations, like fan clubs and, of course, like communities (Kling, 1996). His descriptions of the activities that transpire inside online communities correspond to a rich range of human interaction found in any group situation, from commerce to flirting, but the focus on a conversational forum as the essence of community essentially flattens the concept and strips it of its social complexity.

Rheingold’s definition of virtual communities has been highly influential in subsequent literature, and most published descriptions of online groups quite casually characterize the forum in question as a community, almost by virtue of the fact that its members interact online. Sometimes the use of the term is almost apologetic, in recognition of its market dominance. For example, according to Preece (2000), community is, in the final analysis, the best term to use for online social interaction, given that its “[w]idespread use by e-commerce entrepreneurs has in fact made the term a buzzword” (p. 9), and that it “has also become a blanket term to describe any collection of people who communicate online” (p. 17).

What is most praised by writers about virtual communities appears to be a sense of mutual engagement and openness among members, a “feeling” of community that is not analytical (cf. Bays & Mowbray, 2001; Bird, 1999; Kim, 2000; Rheingold, 1996). There seems to be an underlying equation of the term with solidarity, reciprocity, and support, even when criticism and conflict is acknowledged (Komito, 1998). These traits are very important, but they are equally characteristic of many kinds of groups, clubs, and other social forms that do not call themselves communities. In lay language, “sense of community” is often an idiomatic shorthand for these kinds of feelings, and particularly widespread in the North American cultural context. In short, community has a strong symbolic value that does not necessarily characterize real-life group interaction (Cohen, 1985; Miller, 1999). This implicit understanding of community issues a normative expectation regarding how participants in online forums ought to interact, but provides no clue as to how such interaction can be structured and motivated.

An extreme example of the misconceptions generated by the casual use of the term “virtual community” can be found in a recently published description of a Web site, Through Our Parents’ Eyes: Tucson's Diverse Community (Glogoff, 2001) . The project, an online multimedia resource that brings together a wealth of heretofore unavailable information on the cultural heritage of Tucson residents, has inspired e-mail comments to the site’s author from readers who express their appreciation, and who often share further details about their heritage and how the site has contributed to their self-awareness and pride. These social interactions, in the form of letters written directly to the site’s owner and not published on the site for further discussion, form the basis for what Glogoff calls a “virtual community,” when in fact they are semi-private, bilateral communications. The site offers no support for visitors to identify and communicate with other interested visitors. It is a strange form of community where the participants do not and cannot easily recognize and interact with each other! We are in no sense critical of Glogoff's Tucson project, but rather of the sloppy, romantic way in which he has characterized the kind of social formation that has developed around it:

William J. Mitchell, in his important forward-looking book City of Bits: Space, Place, and the Infobahn (1995), predicted that “being online may soon become a more important mark of community membership than being in residence.” This certainly has been the case with many of the Web exhibits in Through Our Parents' Eyes. They have attracted, for several years, people who self-identify with its content and seek out some degree of membership. Such behavior is consistent with the view that an essential element of building a climate of trust involves “feeling secure in revealing vulnerable parts of ourselves to others.”
It could be argued that a kind of “imagined community” of Tucson is being developed through the Web site, analogous to Anderson’s (1991) discussion of nations as “both inherently limited and sovereign,” in which
the members of even the smallest nation will never know most of their fellow-members, meet them, or even hear of them, yet in the minds of each lives the image of their communion. (p. 6)
Glogoff's uncritical characterizations of social formations as communities do not promote understanding of the different types of social relationships that actually develop in electronic forums. Nor do they help devise ways to support e-forums through design, based on the characteristics of the target population and the theme of the venue. Using the term community to describe as diverse a range of groups as auction bidders, students in a distance education course, investors, and cancer patients (cf. Hagel & Armstrong, 1997; Werry & Mowbray, 2001), is unhelpful to scholars, designers, and even prospective participants. The use of the term is aspirational, and may be largely based on the need for recruitment of members, rather than on a description of empirical reality.

In contrast, Haythornthwaite, Kazner, Robins, and Shoemaker (2000) found that the extent to which the students in an online degree program can be characterized as a community is a question to be resolved empirically, rather than by assumption. Analysis of the interpersonal ties that students developed in a distance education course that began with a face-to-face “boot camp” revealed many traits that are consistent with accepted definitions of community: recognition of members and non-members, a shared history, a common meeting place, commitment to a common purpose, adoption of normative standards of behavior, and emergence of hierarchy and roles. These data, although not identical to Brint’s (2001) findings, would also support a community conception based on his six criteria. Interestingly enough, the study also documents how students eventually “disengaged” from the community as they finished the program.

Learning Communities?
According to Liana Nan Graves (1992), the term “learning communities” became popular among educators in the 1990's. Educators seem to take two major approaches in conceptualizing learning communities. One approach characterizes them in terms of curricular organization, such as “linked courses, which link cohorts of students taking two courses in common with one course typically content-based and the other application-based” (Kellogg, 1999). Kellogg discusses five different models that emphasize the curricular structure for learning communities in higher education. This approach, centered around resource-sharing and structural arrangements, is similar to many existing e-forums, where the arrangement and accessibility of resources through socio-technical design can be central.

Other educators emphasize the social relationships among participants as most critical in conceptualizing learning communities. In particular, they stress certain kinds of human sociality that they believe are desirable, such as high levels of cooperation and collaboration between students, as well as between students and teachers (Graves, 1992). In this sense of the term, Graves uses the term community to signify “an inherently cooperative, cohesive, and self-reflective group entity whose members work on a regular, face-to-face basis toward common goals while respecting a variety of perspectives, values, and life styles.” She itemizes several additional characteristics of learning communities: 1) “where everyone feels they belong and are respected”; 2) “where interaction is on-going, face-to-face, regular, and focused around common goals”; and 3) “a cohesive yet self-reflective group.”

Developing and supporting cooperative learning groups is a major challenge, since the participants are asked to engage in more personal risk-taking behavior than in typical courses. Instructors are encouraged to step down from their stages (pedestals), and to act as coaches and co-participants who can display ignorance as well as knowledge. Social relationships may be multivalent, as when students are supposed to collaborate, but are also graded individually, and thus competitively. Graves (1992) discusses some approaches to developing learning communities which rely on carefully planned stages of activity for encouraging trust and strengthening ties among members. Although a significant amount of work is required by teachers or facilitators, the concept of building a scaffolding for learning communities may be very useful for developing other kinds of communities as well.

Riel and Fulton (2001) write about the ways that e-forums may support the development of geographically distributed learning communities. They write, however, as if face-to-face learning communities are an easily accomplished practice. Their example of a geographically dispersed learning community of teachers emphasizes the technological complexities (and anxieties) that teachers face and the ways that the participating teacher can require significant time to find ways to effectively use electronic media in their own teaching. In contrast with Graves, Riel and Fulton do not discuss how participating in a normative community can be difficult in terms of negotiating identities, learning to trust other participants, and take professional risks (such as frankly acknowledging difficulties and failures).

In our own experience, observation, and reading of empirically reliable research, creating online groups of any form tacitly requires finding ways to support the social processes that would be typical of face-to-face groups, in addition to dealing with the complications of communication in a specific e-forum. When authors such as Riel and Fulton ignore the fragility of the social processes for developing effective learning communities of any kind, they set up unrealistic expectations about the issues that must be engaged in fostering distributed learning communities via e-forums.

A common theme: Building trust
While many people portray groups as caring arrangements in which to share experiences and thoughts, and to learn new ideas and practices, group participation can involve risks —and thus require trust between participants. Students who share good information, but who are graded competitively, risk earning a lower grade than if they did not share. Students who ask for help may be seen as incapable (although that may also be a mark of pride in some schools for some topics). In professional development groups, teachers who share experiences of troubled teaching, and even of failure, risk being viewed as incompetent by their peers. Those who propose a novel idea risk the possibility of being seen as strange and perhaps even ostracized. We say that people are more willing to take risks in groups that develop a high level of trust.

There are several ways to characterize trust, and one of these is as a measure of risk. For example, Mayer, Davis & Schoorman (1995) define trust as “the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action, irrespective of the ability to monitor or control that other party” (cited in Grabowski & Roberts, 1998). We can interpret each of the activities in the previous paragraph through the lens of this definition. Trust can develop over time, as people reciprocate in sharing information, demonstrate respect for one another, are careful in maintaining certain information as confidential, and so on. Each of these steps involves a certain measure of risk, and when that risk is rewarded, trust is more likely to develop.

While trusting relationships can facilitate learning in groups, they also facilitate the practical work of a wide variety of organizations. Thus “trust at work” has been the subject of notable research in the field of organization studies (Kramer & Tyler, 1996). Of course, not all workplaces exemplify high levels of trust. For example, it is common for telephone operators and the staff of “800 number” service call centers to risk having any telephone call monitored, and may have their productivity measured daily. In contrast, the participants of engineering design teams must collaborate for months or even years before they have a “product” that can be assessed.

Working online complicates the formation of sustainable trust between people. In a face-to-face setting, people can see each other's appearance, and gauge some emotional reactions during a conversation. Appearances can be deceiving, or simply lead to stereotyping. But we are sufficiently accustomed to living with appearances that we all develop ways to interpret them for developing trust. In online communication, the participants have to work hard (through writing, today) to communicate something about themselves, and about each of their reactions. Grabowski and Roberts (1998) note that "[d]eveloping trust in VO's [virtual organizations] requires constant, continual communication among members to build relationships that provide the foundation for trust.”

Jarvenpaa and Leidner (1998) studied the behavior of approximately 70 virtual teams composed of students in different countries enrolled in masters degree programs. They noted that e-forums offer the possibility for teams of people to work in different countries and collaborate. But they noted that “a dark side to the new form also exists: such dysfunctions as low individual commitment, role overload, role ambiguity, absenteeism, and social loafing may be exaggerated.” They collected systematic data about these teams, and selected 12 for detailed investigation. In particular, they were interested in teams that started with low trust and those that started with high trust, and that also ended with either low trust or high trust. They examined “trust trajectories” over the lives of the teams. They found that some behaviors facilitated trust early in a group's life (i.e., communication of enthusiasm, coping with technical uncertainty). However, other behaviors helped to maintain or enhance trust during a team’s activities (i.e., predictable communication, substantial and timely responses).

Jarvenpaa and Leidner's study is important reading for anyone who is trying to develop learning communities online. Their specific findings may or may not translate literally to other settings. But some key ideas stand out: the importance of trust in teamwork; that groups may begin with varying levels of trust; and that the groups that increase their levels of trust or maintain high levels of trust work hard to do so in ways that are describable. In short, group solidarity online in the service of anything —work or play— is a fragile accomplishment rather than a gift that just “comes with the territory.”

Some educators have found that it is easier to develop the bases for trust in online groups through face to face means, and then to continue group work online. An interesting example, mentioned earlier, is Haythornthwaite et al.’s (2000) account of an online masters degree program at the University of Illinois. Each new cohort of students is required to visit the Urbana-Champaign campus to study together for a week in a “boot camp” at the beginning of each academic year. The students learn how to use various relevant technologies at “boot camp.” But they also work in varied groups, and the workload is such that high levels of cooperation are required. At “boot camp,” these students come to know and trust others in the class. Haythornthwaite et al.'s interviews with students during their work in the program indicates that these ties are exceptionally helpful after the students disperse and continue their study online. We realize that a practice that may be workable for an elite graduate program may not be feasible or even meaningful for other online courses and study groups. But it anchors the point that trust does not automatically develop for many groups online; it may require significant intervention for e-forum organizers to foster trust, either online or face-to-face, or both.

Participating in the e-ILF

Initial Conceptions and Aspirations for the e-ILF
The Inquiry Learning Forum (ILF) was created in 1999 by researchers at the Indiana University School of Education as an online forum to support inquiry-based teaching and learning practices among Indiana science and math teachers at the secondary school level (see http://ilf.crlt.indiana.edu). The underlying conception was to provide a set of interesting teaching materials and online forums, the e-ILF, where teachers could discuss them and reflect on their own teaching practices, and function as a “community of practice” (Barab et al., in press).

“Communities of practice are informal networks that support professional practitioners to develop a shared meaning and engage in knowledge building among the members” (Hara, 2000, p. 11). This definition is based on the Wenger's elements of communities of practice (CoPs): negotiating meaning between participants; preserving and creating knowledge; and supporting the development of identities (Wenger, 1998). Some recent studies of how people learn technical, craft, and professional skills found that new entrants to an occupation learn informally through discussions and observations with others who do similar work. These studies include a wide range of occupations, including flute-makers, office equipment technicians, and lawyers in public defenders’ offices.

The ILF research team was concerned that in-service teachers have limited opportunities to discuss and observe teaching practices with other teachers because they are often all in their own classrooms at the same time. In contrast, for example, Hara (2000) found that the public defenders in a county court would attend each others’ major trials to observe how their co-workers interacted with juries, judges, and witnesses. Being a good public defender is not simply based on knowing the law, or in developing the basis for a case. A public defender has to work with witnesses and evidence, in front of a judge, to convince a jury of a particular position. In a similar way, being a great math teacher is not based simply on knowing a great deal of mathematics. Great teachers develop strategies and a language for communicating mathematical concepts to their students and stimulating their interests in mathematics. However, teachers have little opportunity to visit each others’ classrooms to observe and to discuss their teaching in action.

In 1999, the School of Education faculty who were associated with the ILF interviewed a number of science and mathematics teachers about their professional development experiences (see MaKinster, Barab and Keating, 2001; Barab et al., in press). They found that many of their in-service teacher informants expressed an interest in seeing other teachers teach. These faculty translated these interests into a conception in which an e-ILF would provide facilities that would provide some of the key elements of a classroom visit, such as videos of class segments, the relevant lesson plans, and reflections by the teacher about his or her class. In addition, each virtual classroom would support a local discussion list in which teachers could reflect on the class materials and on their own teaching. The Education faculty expected communities of practice that focused on inquiry learning in math and in science to develop through these online discussions. Other features of the e-ILF include a lounge, for more general threaded online conversation, and thus further support for the development of CoPs, as well as spaces for closed working groups and a resource center. (See Barab, MaKinster, and Scheckler, this volume, for a more detailed discussion of the e-ILF's design.)

The ILF researchers took a strong stance that fostering inquiry-oriented teaching by enhancing teachers’ understanding of their own classroom practice was a primary focus of the site. The research team knew that certain services, such as providing a large library of “canned lesson plans” that teachers could download and apply immediately, would almost certainly attract many more teachers. But they did not see such a “lesson plan service” as enhancing teachers’ capabilities in the long run. They saw it as more akin to enabling students to download term papers, rather than developing the skills to write their own. Thus, the lesson plans on the e-ILF were part of a larger portfolio of materials in a specific classroom that illustrated teaching a specific topic (such as the concept of temperature or Pythagorus' theorem).

In 2000, the ILF staff identified 14 Indiana teachers who were willing to have parts of their classroom activities videotaped, and posted in an online “classroom” about their teaching on the e-ILF. In addition, these teachers provided their lesson plans and notes about teaching this lesson, that were also posted in their e-ILF virtual classrooms. We estimate that each teacher spent about 30 hours working with the ILF staff to develop a virtual classroom that was organized around one lesson.

The ILF research team expected teachers to develop reflective online discussion groups on their own. In fact, one major research question was, “How do the ILF members structure themselves into communities and how do we promote boundary crossing?” Barab (1998) characterized communities of practice in these terms: “Much like a living organism, they are self-organizing, and cannot be designed prima facie. They grow, evolve, and change dynamically, transcending any particular member and outliving any particular task.”
Some structures, such as the sample classrooms and a lounge, were created by the ILF staff. The ILF staff worked with specific teachers to develop materials for each online classroom. But the ILF staff were not expected to moderate any of the resulting discussions. There were also few cues and clues for participants about how they should behave in the e-ILF, or what specific expectations they could have about others' participation (i.e., frequency of posting). This model of autonomous group organization has been very successful for such professional e-forums as scientific Usenet Newsgroups, Bionet, and non-recreational services such as the message boards on ParentsPlace.com.

The registered e-ILF members include in-service and pre-service teachers, along with ILF researchers and developers. Those who wish to participate in the e-ILF are screened by ILF staff to ensure that they are either appropriate teachers, student teachers, or associates of the project. Potential participants such as parents and journalists are not accorded access to the e-ILF, by request of the teachers who helped to create its virtual classrooms.  The participants on the e-ILF are identified by pseudonyms they choose themselves (often compressed versions of real names), although ILF researchers and highly active teachers include their real names in their postings and profiles. ILF staff have access to the identities and profiles of all members, who in turn can control the level of personal and professional information revealed to other members on the site.

 e-ILF Activity from Spring 2000 to Mid-2001
By late 2001, there were 21 videotaped “classrooms” on the ILF Web site. Over 300 in-service teachers out of approximately 26,000 secondary math and science teachers in Indiana had registered for the ILF, along with some 430 pre-service teachers at Indiana University.

However, participation in the e-ILF has not reflected the breadth of this membership: statistics collected in February 2001 show that only 14% of in-service teachers had visited the site five or more times, and 9% had visited ten or more times. Online discussion by in-service teachers has been limited to a relatively small group, whose numbers have increased rather slowly: 58 of the site’s 320 registered in-service teachers (18%) had posted a total of five or more messages by October 2001, up from 35 of 202 (17%) by February. Only 6% had posted a total of 10 or more times by October. Pre-service teachers have recently begun posting more actively on the site, principally because e-ILF participation is often part of their learning requirements: 147 of 432 pre-service teachers (34%) had posted five or more times by October 2001, up sharply from 51 of 183 (28%) before February.

The number of posted messages connected to each “classroom” varies widely, but a typical set of 15 posts is threaded into 6-7 topics with anywhere from 0-3 responses each. The “lounge” section of the Web site contains general topics for discussion, with a typical topic taking up 15-30 posts over a one-year period.

These levels of participation are not unusual, particularly for such a new forum. Empirical research on other e-forums also shows evidence of highly concentrated levels of participation. For example, Nonnecke and Preece (2000) found that over 70% of subscribers to health-related discussion groups posted three or fewer times in a 12-week period; lurking levels for software-related discussion groups exceeded 90%. Selwyn (2000), in his analysis of two years of postings to an electronic discussion group for teachers in Great Britain, found that one-third of all posted messages came from only 26 of its 900 members.

But ILF research team members have expressed concern that the online exchanges among teachers do not display the levels of engagement and critical reflection that were initially expected from what was to be a community of learners. An example of a typical supportive post:

Just watched your video. It's good to see someone brave enough to go on film. You seem to enjoy what you are doing and your class seems to show you much respect. I think it is great to see a teacher laugh with her students and they laugh back. Great job!
The videotaped classrooms nevertheless represent a touchstone for discussion and learning. Many teachers’ posts are requests for further information that could help them link their own methods to the practices observed on the video clips:
I think you had a good idea of using 2 people in a group and having two questions (one for each to do) and one to do together. What do you do if you have a student who doesn't want to participate at all?
In the “lounge” section of the e-ILF, teachers discuss substantive issues that arise in math and science teaching, and present many different views on inquiry-based teaching and learning. In this area, the teachers’ posts consist of opinions or questions without reference to the videotaped classrooms, and they often mention useful outside resources. Although there is little evidence of direct critical engagement with other opinions, and many postings go unanswered, the tone of the postings is self-reflective, helpful, and engaging. For example:
…Many of my students lack this background knowledge and I am at a loss on how to teach the standards that are expected, encourage inquiry and supply sufficient background knowledge.

…I think the important thing to do is to find out what your students know and go from there. If your students are struggling with open or guided inquiry, go back and try something more concrete.

…What does everyone feel about that? Have you experienced problems with open inquiry in advanced or basic classes?

In response to the perceived lack of critical engagement, some ILF staff have posted some messages to stimulate more engagement, by asking questions in response to comments, making critical observations, and referring to specific aspects of the videos. For example,
Can you be more specific. What subject? What techniques?

…I would have liked to see your lecture extended and given as a separate lesson. You were covering so much relevant and necessary information. I felt like you were trying to hurry through information and especially at first you were simply feeding the information, instead of trying to elicit it from the students…


According to Selwyn (2000), one indicator of community is an explicit sense of group identity that some participants mention in their discussions. In his study of teachers’ use of an electronic discussion list, Selwyn found little if any evidence of such a “sense of community,” of engagement among members in a common mission. Likewise, in the e-ILF, the messages posted in both the “classroom” and the “lounge” forums lack any mention of the group. The postings tend to be about general and specific aspects of the forum’s common idea, inquiry teaching and learning, but do not express a bond among its members as Indiana teachers engaged in a common mission or as e-ILF participants. More importantly, Selwyn also found that participants carefully maintained a professional and formal air as teachers, taking care not to risk exposure as incompetent. This appears to be replicated in the e-ILF. As in Selwyn’s group, there is no “staff-room talk” on the e-ILF that could build trust and help forge closer personal relationships among members, thereby adding another dimension of community.

The e-ILF might be contrasted, for example, to an online financial site visited over a one-year period by one of the authors. The site was organized into dozens of individual discussion groups. In a few of these, participants gradually revealed additional information regarding their identities if they felt confidence in the group’s usefulness and in the reliability of other members; conversely, new members were often asked to reveal personal information or “bona fides” in order to obtain a thoughtful response to a question. Moreover, in those groups in which mutual confidence was rewarded, there were numerous references to group identity and a common mission, and members often engaged in other forms of social interaction, such as migrating together to new groups, meeting off-line, or interspersing philosophical and personal discussions. Some participants criticized others harshly, although the more cohesive groups tended to exercise peer pressure to maintain norms of civility and encourage open, trusting exchanges.

Thus, with the exception of commitment to the common idea of inquiry-based teaching, the e-ILF does not correspond to the sociological criteria for community identified by Brint (see above). Nor does it satisfy Graves' aspirational criteria for learning communities. Nevertheless, some teachers are using the e-ILF to learn and share new ideas among peers, and their numbers are steadily growing. In this sense, the site is a valuable peer-to-peer resource center for in-service and pre-service teachers to view teaching examples, obtain ideas for their own practice, and share opinions on the subject of inquiry-based teaching and learning. As such, the continuation of the e-ILF along present lines suggests an online social formation equivalent to a useful drop-in center for teachers’ professional development in the area of inquiry-based teaching. The ILF could provide important benefits to its members and their students, but the activity is still in a very early stage, and it is too soon to assess productively.

Even in face-to-face situations, promoting both collegiality and critical engagement among teachers has long been perceived as a problem. In fact, the nature of the ILF interactions described above are consistent with empirical research on teachers’ interactions in general, which are characterized by a lack of direct advice or criticism (Ellis, 1993; Little, 1985, 1990). Overcoming this “etiquette” (Little, 1985) and promoting greater critical engagement and sense of group identity might require mechanisms —both online and offline— conducive to forging greater trust among ILF members for open, frank discussions and engagement. If the e-ILF aspires to move beyond its present resemblance to a professional development site in order to build stronger mutual ties among members that more resemble a community, certain structural changes might be necessary to stimulate such engagement, trust, and group identity.

Building support for community activity in the e-ILF
In her discussion of learning communities, Graves (1992) notes:

In the early stages of building a community —whether classroom, school, faculty group, or administrative team— people are concerned with finding a place for themselves within the group. This involves:

• becoming acquainted with other group members on a friendly basis: Who are you?
• presenting oneself to the group and being accepted as a valuable member: Who am I? ...

.... classrooms, schools, and educator support groups are not natural communities — that is, there is usually no built-in reason such as kinship or generational village ties to bring these people together. ... We need to encourage strong standards of equality and honoring of diversity ... We can do this in two ways. The first is by offering opportunities to get to know one another under circumstances that maximize enjoyable interaction while creating opportunities for information exchange and problem solving.

Graves goes on to discuss a wide variety of group-building activities which may help people come to know each other and work together under differing levels of risk taking and accomplishment. Many of these observations are applicable to the challenges facing the e-ILF. Here, building community among its members will probably require structural changes to provide both spaces and mechanisms in which trust, familiarity, and group identity can be forged among members. In keeping with our conceptual understanding of online spaces as both socially and technically structured, we believe that such changes cannot only be instituted through technological redesign, but must include offline measures as well.

The expectation that teachers would self-organize CoPs in the e-ILF’s virtual classrooms has not yet been realized. In light of Graves’ observations, we suspect that the development of CoPs could require a much more interventionist strategy in which moderators try to encourage participants to get to know each other (for example, by posting some personal information as well as professional information in their online bios) and help to focus and deepen the online discussions. These kinds of interventionist activities are difficult because they make more demands for participants for self-revelation, and for reading and posting in specific time frames, while the ILF strategy has been to attract participants by making no demands upon them after they register.

Bounded Groups
During the first year of the e-ILF’s operation, there were several requests to support the discussions of groups that wanted private spaces. One such group, that is developing a curricular framework for mathematics (Collaboration to Enhance Mathematics Instruction, CEMI), is led by a key faculty member of the ILF research team. The workspaces for the bounded groups include private discussion lists, as well as spaces to post documents, link to external resources, etc.

This request for a space on the e-ILF was accepted, although the research team was initially reluctant to encourage more bounded groups. Subsequently, several additional groups requested bounded spaces for their own discussions. These included a group of high school physics teachers and a physics professor who are developing a collaborative experimental project, a group of biology teachers who are developing curricular plans to study watersheds, and another group of biology teachers who use salamanders in their classes. After continuing deliberations, the ILF team decided to support such groups and encourage others.

As of early 2002, over two dozen bounded groups have been formed on the e-ILF. These groups constitute naturally forming teams and potential communities of practice, in which relations of trust among members might be expected to develop more easily than in the open forums, for reasons we have explained above. The hard work of team or community development —including trust-building, and of recruiting participants— is done by others who are usually based outside of the ILF. The e-ILF has the relatively easy task of providing some of the communications infrastructure to enable these groups of teachers to carry out their work, although it is likely that the ILF will also have to play a role in encouraging their formation and stimulating both online and offline activities. It remains to be seen whether the growth of interactions within bounded groups will also eventually lead to interactions across them, a development that could strengthen the ILF in a way that is consistent with its designers' goals.

Ironically, the importance of bounding groups to facilitate trust in electronic forums has been known before the development of the ILF project. Feenberg (1986) identified its importance for electronic conferences and Bakardjieva, and Feenberg  (2002) note that it is an  important feature to support democratic relationships in electronic forums. DiMauro and Gal (1994) recommended that online forums have for bounded groups to support trust between teachers. Somehow, these ideas did not travel rapidly enough to Bloomington!

The e-ILF in Comparative Perspective

Reflective dialogue and trusting group behavior online
We began this chapter with a provocative stance, by claiming that “the casual use of community to characterize groups that are engaged in learning, or groups that participate in e-forums, is seriously misguided.” The ILF project has a set of ambitious goals that rest on the possibility of developing online forums that enable teachers to improve their abilities to teach in an inquiry style through participating in communities of practice. This approach rests on several complex assumptions: (a) that a critical mass of teachers would be willing to participate actively in some kind of online forums about inquiry-learning; (b) that open-ended discussions of sample teaching episodes would be valuable to participating teachers; (c) that teachers would self-initiate reflective dialogue about the online classroom materials and their own teaching; (d) that teachers who were participating in the e-ILF would develop a sense of group identity and forge mutual trust via online interactions; and (e) that their group identities would be transformed over time into lively communities of practice.

Translating each of these assumptions into a practice on the e-ILF is a significant accomplishment. As Graves noted, developing a community is an accomplishment. Unfortunately, in the early stages of the ILF Project, the tendency to characterize the e-ILF as a community space distracted attention from the vexing issues about what would be required to transform it from an electronic forum with facilities that could support conversations into a space in which communities of practice were actually forming.

Our examination of interactions on e-ILF shows that although it has value as an online resource for teachers, it does not at this time resemble a community. It is easier to provide communication infrastructure for pre-existing groups than it is to use an e-forum to develop high-performing groups and CoPs directly. Thus, the e-ILF may benefit from the addition of bounded groups, that involve greater engagement, stimulate trust, and are complemented by off-line activities, thereby enabling the ILF to support a wider variety of inquiry projects. At the same time, the ILF could play a less demanding social role of having to support all stages of development of the community of practice it envisions.

Unfortunately, the difficulties of supporting reflective dialogue and community building in online forums seems to be under-appreciated within the educational communities. The best research that we have found about trust building for online groups has been conducted by Information Systems faculty who study teamwork online (see above). In contrast, Sherry (2000) is all too typical in unreflectively reporting claims that:

One advantage of text-based communication is that written communication tends to be more reflective than spoken interaction. ‘The very act of assembling one's thoughts and articulating them in writing for a conference audience appears to involve deeper cognitive processing’ (Berge, 1997, p. 10).
Claims framed in this way make it appear that the use of text-based electronic communication will almost certainly lead to reflective dialogues where assumptions are evaluated, alternatives discussed, contexts carefully explored, etc. Yet, according to research on typical teaching practices in the U.S., teachers rarely engage in this kind of discussion when they are meeting face-to-face (Little, 1982, 1985, 1990). Perhaps we should not be surprised to find that such reflective discussion does not spontaneously arise in the e-ILF.

Our observations of the e-ILF discussions are consistent with mid-1990s research findings from the LabNet project, an e-forum to support the professional development of science teachers. DiMauro and Gal (1994) examined the online discussions of a group of science teacher leaders who were acting as liaisons to the staff of an online professional development project. A network infrastructure was designed to enable these teachers to “exchange to reflect upon their involvement with peer leadership and teacher-teacher support.” They characterized messages as informative, responsive (to a query), or reflective (one in which a participant “thinks out loud” about some teaching practice and different ways of approaching it). DiMauro and Gal note that reflective postings were very infrequent. Most seriously, they observe, “Reflective responses are difficult to formulate and risky to post because of the personal nature of the content.”

They speculate about the conditions that support reflective postings, and suggest that they include “protected workspace for reflection, retrieved text base, collaborative research, access and response to messages, structure dialogue, linking action with reflection, forming reflective practice inquiry, and participatory motivation.” Some of these conditions are found on the e-ILF (i.e., retrieved text base). However, only the bounded groups have a protected workspace for reflection. In contrast, postings in the virtual classrooms may be read by any student teacher or in-service teacher who joins the e-ILF in the future and, as a result of recent changes to the site, by school administrators.

In a related study that is also based on observations of LabNet dialogues, Spitzer, Wedding and DiMauro (1995) published a significant set of suggestions for promoting more reflective discussions on LabNet. Apparently, the LabNet discussion groups were usually moderated. Like Graves, they recommended that the forum moderators explicitly find ways to build trust among participants, such as “get to know each other by talking about your situation, interests, etc.” They tried to promote group development with suggestions such as, “Invite people to join via e-mail and e-mail lists”; “Move a private e-mail dialogue to a public forum”; and “Ask thought-provoking questions that ask for another's idea or point of view, and make your intentions explicit.” Their strategies, which encourage the moderators to engage members in thinking and doing, are extremely labor-intensive. Unfortunately, they provide no data about how well various LabNet moderators adopted these strategies (or other interventions), and with what effects. Nonetheless, Spitzer et al. suggest that supporting reflective dialogues online and developing community requires significant and committed communicative and social work.

In the end, community development is likely to be a complex accomplishment that is difficult to initiate without purposive interventions from some kind of leaders or stewards. It will rarely happen online alone through self-organizing.

How do lively online places happen?
This discussion leaves open the question about how some professional e-forums as scientific Usenet newsgroups, Bionet, and non-recreational services such as the message boards on ParentsPlace.com, become lively places. In our view, they demand less from their participants than does the e-ILF. Most of the sites like these that we know are organized as professional or personal help groups. For example, the postings that we have read on BioNet are often focused on questions raised by a stymied researcher who has trouble getting some specific piece of equipment to work.

For example, a posting in a forum for discussing ACEDB (A C.Elegans database) in June 2001, was framed as follows:

Hello, I am trying to set up saceserver on my machine, following the "User Guide To Sockets-based Client/Server"; I experience some troubles.

I am using acedb 4.9a. saceserver (and saceclient) are doing just fine when launched in the foreground. But when launched through inetd (actually, it is xinetd on my machine), I get a FATAL ERROR - (fc on grun.marseille.inserm.fr) reported by program saceclient (ACEDB 4_9a), in file acesocketlib.c, at line 284: connect error system error 111 - Connection refused

/etc/services :
acedb 20113/tcp

/etc/xinetd.conf :
acedb stream tcp wait acedb /usr/local/bin/saceserver saceserver
/database/acedb 200:200:0

acedb is a valid user on my machine

saceserver is -rwxr-xr-x

Any help will be appreciated!

A response one week later resulted in a technical dialogue that lasted another two weeks. The form of dialogue, “question and answer” is also common for teachers (Lewison, 2001). It is similar to the informative postings identified by DiMauro and Gal (1994) on LabNet, and by Selwyn (2000) on a teachers’ discussion list. We have seen a significant amount of highly focused technical Q&A's on the BioNet groups, and no significant discussions that sound like scientific reflections. Some of the lists on BioNet support researchers who participate in a tightly knit research community and who have numerous complex interactions off-line as colleagues, in conferences, as editors and reviewers, and so on. (In this sense, in fact, BioNet resembles some aspects of the e-ILF's bounded groups.)

In another realm, www.ParentsPlace.com offers its participants an opportunity to raise questions about parenting and have a good chance of hearing comments from other parents. The primary dialogic format is also “question and answer.” Further, most participants are pseudonymous. One recent paragraph-long post about toddlers began:

Hi everybody, I don't usually post here but.....desperate times call for desperate measures! I babysit for four kids, and the boy who's 3.5 has had some aaawful tantrums lately and I have no idea what to do about it.
The writer received a pseudonymous reply asking some questions and offering some advice within a day. Most of the dialogues on ParerentsPlace.com seem to be Q&A; few comments are reflective. Most seriously, the pseudonymous writer is able to express extreme desperation based on her inability to manage her relationship with the 3-1/2 year old boy. Yet based on the experience of the teachers’ online forum studied by Selwyn (2000), it would be implausible for an e-ILF participant to express this level of interpersonal incapability (and thus professional vulnerability) in discussing dilemmas of teaching. In terms of willingness to risk looking incompetent, there appears to be a significant difference between the situation of entirely pseudonymous parents and caregivers, on the one hand, and that of teachers —whose anonymity is not fully ensured by the e-ILF, for example— on the other. Yet in order to build the type of community of practice sought by the ILF, teachers must become willing to engage in professionally risky conversations in order to build trust and group identity. The overall level of support work required by its moderators and organizers, therefore, will be concomitantly greater.

Conclusions
The ILF has ambitious goals in encouraging both reflective discussions of teaching via inquiry and also supporting the development of communities of practice in support of inquiry teaching. To fulfill these goals, the ILF must be capable of supporting a high level of trust-building and other social practices that cannot be readily “wired into” the design of an e-forum, but instead require a significant investment of time and human resources, and perhaps even offline work as well. It is possible that some arrangements that offload the community building to others, such as the bounded groups, may enable the e-ILF to provide a different kind of useful role in supporting professional development activities of geographically dispersed teachers.

We have illustrated two approaches to building online groups that differ sharply. The original ILF strategy of bring teachers together via discussions of classrooms illustrates “IT-led group development.” In contrast, the bounded groups illustrate “IT-supported group development.” The IT-led strategies are much more difficult to make workable. Some of the best examples that we know of, such as Parentsplace.com, do not place a strong emphasis upon group formation, but instead function essentially as peer-run resource and mutual help centers. Other examples, such as online group game sites, offer their players a form of entertainment by participating in game-playing “guilds.” Yet the expectation of using IT to play the leading role in forming close, trusting groups is not likely to be fulfilled. In contrast, effective IT-supported groups are very common, since they do not require that the various and complex processes of group formation and group development rely principally on an electronic forum. Instead, the role of the e-forum is to enhance, extend, and support wider group processes and goals. We believe that this approach will be more fruitful, not only for the ILF, but also for a wide range of professional, group, and learning endeavors.

Acknowledgements
This research was supported in part by a National Science Foundation Grants, REC-9980081 and 9872961. We have also benefited from discussions about professional development with Cathy Brown and Mitzi Lewison. Blaise Cronin, Andrew Feenberg, Rebecca Scheckler and Murali Venkatesh also provided helpful comments.

References
Ahuna, C. (2001). Online game communities are social in nature. Retrieved July 15, 2001, from the World Wide Web: http://switch.sjsu.edu/v7n1/articles/cindy02.html

Anderson, B. R. (1991). Imagined communities : Reflections on the origin and spread of nationalism (Rev. and extended ed.). London: Verso.

Bakardjieva, M. and A. Feenberg. 2002. Community Technology and Democratic Rationalization. The Information Society 18(3):181-192.

Barab, S. A., & Duffy, T. M. (2000). From practice fields to communities of practice. In D. Jonassen & S. Land (Eds.), Theoretical Foundations of Learning Environments (pp. 25-56). Mahwah, NJ: Lawrence Erlbaum Associates.

Barab, S. A. (1998). Theoretical aspirations: Communities of practice.

Barab, S. A., MaKinster, J., & Scheckler, R. (in press). Designing system-dualities: Building online community. (This volume)

Barab, S. A., MaKinster, J., Moore, J., Cunningham, D. & the ILF Design Team. (in press). The Inquiry Learning Forum: A new model for online professional development. To appear in Educational Technology Research and Development.

Bays, H. & Mowbray, M. (2001). Cookies, gift-giving, and online communities. In C. Werry & M. Mowbray (Eds.), Online communities: Commerce, community action, and the virtual university (pp. 47-70). Upper Saddle River, NJ: Prentice Hall.

Berge, Z. (1997). Computer conferencing and the on-line classroom. International Journal of Educational Telecommunications, 3(1), 3-21.

Bird, S. E. (1999). Chatting on Cynthia's porch: Creating community in an e-mail fan group. Southern Communication Journal, 65(1), 49-65.

Brint, S. (2001). Gemeinschaft revisited: A critique and reconstruction of the community concept. Sociological Theory, 19(1), 1-23.
Cerf, V. (2001, May 15). A brief history of the Internet and related networks. Internet Society. Retrieved June 22, 2001, from the World Wide Web: http://www.isoc.org/internet-history/

Cipollone, T. (1995). The plain talk planning year: Mobilizing communities to change. Baltimore, MD: The Annie E. Casey Foundation.

Cohen, A. P. (1985). The symbolic construction of community. London: Tavistock Publications.

DiMauro, V., & Gal, S. (1994). The use of telecommunications for reflective discourse of science teacher leaders. Journal of Science
Education and Technology, 3(2): Retrieved October 2, 2001, from the World Wide Web: http://www.terc.edu/papers/labnet/Articles/Reflective/reflective.html

Ellis, N. E. (1993). Collegiality from the teacher's perspective: Social contexts for professional development. Action in Teacher Education, 15(1), 42-48.

Feenberg, A. 1986. Network design: An operating manual for computer conferencing. IEEE Transactions on Professional Communications, PC29(1) pp. 2-7, March.

Glogoff, S. (2001). Virtual connections: Community bonding on the net. First Monday, 6(3), Retrieved March 7, 2001, from the World Wide Web: http://firstmonday.org/issues/issue6_3/glogoff/

Grabowski, M. & Roberts, K. H. (1998). Risk mitigation in virtual organizations. Journal Of Computer-Mediated Communication, 3(4), Retrieved July 18, 2001, from the World Wide Web: http://www.ascusc.org/jcmc/vol3/issue4/grabowski.html

Graves, L. N. (1992). Cooperative learning communities: Context for a new vision of education and society. Journal of Education, 174(2), 57-79.

Grossman, P., Wineburg, S., & Woolworth, S. (2001). What Makes Teacher Community Different from a Gathering of Teachers?
Seattle: Center for the Study of Teaching and Policy, University of Washington. Retrieved October 2, 2001, from the World Wide Web: http://depts.washington.edu/ctpmail/Publications/PDF_versions/Community-GWW-01-2001.pdf

Hagel, J. & Armstrong, A. G. (1997). Net gain: Expanding markets through virtual communities. Boston, MA: Harvard Business School Press.

Hara, N. (2000). Social construction of knowledge in professional communities of practice: Tales in courtrooms. Unpublished doctoral dissertation, Indiana University, Bloomington, IN.

Haythornthwaite, C., Kazmer, M. M., Robins, J., & Shoemaker, S. (2000). Community development among distance learners: Temporal and technological dimensions. Journal of Computer-Mediated Communication, 6(1), Retrieved Oct 2, 2000, from the World Wide Web: http://www.ascusc.org/jcmc/vol6/issue1/haythornthwaite.html

Hillery, G. A. (1955). Definitions of community: Areas of agreement. Rural Sociology, 20, 111-123.
Jarvenpaa, S. L. & Leidner, D. E. (1998). Communication and trust in global virtual teams. Journal of Computer-Mediated Communication, 3(4), Retrieved July 11, 2001, from the World Wide Web: http://www.ascusc.org/jcmc/vol3/issue4/jarvenpaa.html

Kellogg, K. (1999). Learning Communities (EDO-HE-1999-1). Washington, DC: ERIC Digest.

Kim, A. J. (2000). Community building on the Web: Secret strategies for successful online communities. Berkeley, CA: Peachpit Press.

Kling, R. (1996). Social relationships in electronic forums: Hangouts, Salons, Workplaces and Communities. In  R. Kling (Ed.) Computerization and controversy: Value conflicts and social choice. (2nd edition) San Diego: Academic Press.

Kling, R. (2000). Learning about information technologies and social change: The contribution of social informatics. The Information Society, 16(3), 217-232.

Kolbert, E. (2001, May 28). Pimps and dragons. The New Yorker, 88-98.

Komito, L. (1998). The Net as a foraging society: Flexible communities. The Information Society, 14, 97-106.

Kramer, R. M. & Tyler, T. R. (Eds.). (1996). Trust in organizations. Thousand Oaks, CA: Sage.

Lewison, M. (1997). Writing became a chore like the laundry: The problems and potential of using journals to encourage a reflective approach to teaching. The Professional Educator, 19(2), 13-31.

Lewison, M. 2001. Personal communication, Bloomington, IN

Little, J.W. (1982). Norms of collegiality and experimentation: workplace conditions of school success. American Educational Research Journal, 19, 325-340

Little, J. W. (1985, November). Teachers as teacher advisors: The delicacy of collegial leadership. Educational Leadership, 34-36.

Little, J. W. (1990). The persistence of privacy: Autonomy and initiative in teachers' professional relations. Teachers College Record, 91(4), 509-536.

MaKinster, J. G., Barab, S. A., & Keating, T. M. (2001). Design and implementation of an on-line professional development community: A project-based learning approach in a graduate seminar. Electronic Journal of Science Education, 5(3). Retrieved October 2, 2001, from the World Wide Web: http://unr.edu/homepage/crowther/ejse/ejsev5n3.html

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20, 709-734.

Miller, L. J. (1999). Shopping for community: The transformation of the bookstore into a vital community institution. Media Culture and Society, 21(3), 385-407.

Morris, E. W. (1996). Community in theory and practice: A framework for intellectual renewal. Journal of Planning Literature, 11(1), 127-150.

Nonnecke, B. & Preece, J. (2000, April). Lurker demographics: Counting the silent. Paper presented at the CHI 2000 Conference on
Human Factors in Computing Systems, The Hague, Netherlands.

Oldenburg, R. (1999). The great good place: Cafés, coffee shops, bookstores, bars, hair salons, and other hangouts at the heart of a community (2nd ed.). New York: Marlowe.

Preece, J. (2000). Online communities: Designing usability and supporting sociability. New York: John Wiley & Sons.

Putnam, R. D. (1996, December 1). The strange disappearance of civic America. The American Prospect, 7, Retrieved April 13, 2001, from the World Wide Web: http://www.prospect.org/print/V7/24/putnam-r.html

Quittner, J. (1994, May). The war between alt.tasteless and rec.pets.cats. Wired, 2.05, Retrieved July 11, 2001, from the World Wide Web: http://www.wired.com/wired/archive/2.05/alt.tasteless_pr.html

Rheingold, H. (1993). The virtual community: Homesteading on the electronic frontier. Reading, MA: Addison-Wesley Pub. Co.

Rheingold, H. (1996). A slice of my life in my virtual community. In P. Ludlow (Ed.), High noon on the electronic frontier: Conceptual issues in cyberspace (pp. 413-436). Cambridge, MA: MIT Press.

Riel, M. & Fulton, K. (2001, March). The role of technology in supporting learning communities. Phi Delta Kappan, 82, 518-523.

Selwyn, N. (2000). Creating a 'connected' community? Teachers' use of an electronic discussion group. Teachers College Record, 102(4), 750-778.

Sherry, L. (2000). The nature and purpose of online conversations: A brief synthesis of current research. International Journal of Educational Telecommunications, 6(1), 19-52.

Sills, D. L. (1968). Voluntary associations: Sociological aspects. In D. L. Sills (Ed.), International encyclopedia of the social sciences (Vol. 16, pp. 362-379). New York: Macmillan.

Spitzer, W., Wedding, K., & DiMauro, V. (1995). “Fostering Reflective Dialogues for Teacher Professional Development.” TERC. Retrieved October 2, 2001, from the World Wide Web: http://www.terc.edu/papers/labnet/Guide/Fostering_Refl_Dialogues.html

Sterling, B. (1993). Short history of the Internet. Retrieved July 19, 2001, from the World Wide Web: http://w3.aces.uiuc.edu/AIM/scale/nethistory.html

Tönnies, F. (1887/1955). Community and association. London: Routledge & Kegan Paul.

Weber, M. (1949). The methodology of the social sciences. Glencoe, IL: Free Press.

Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. Cambridge: Cambridge University Press.

Werry, C. & Mowbray, M. (Eds.). (2001). Online communities: Commerce, community action, and the virtual university. Upper Saddle River, NJ: Prentice Hall.

Zakon, R. H. (2001, April 15). Hobbes' Internet Timeline v5.3. Retrieved June 22, 2001, from the World Wide Web: http://www.zakon.org/robert/internet/timeline/