I had also moved on to a more complex mode of fieldwork known as participant observation, and I was getting an education I hadn't expected. Their experience of the world, their ethical sense, the ways they interpreted concepts like work and play were becoming part of my own experience.
Allucquere Rosanne Stone [133]

At the beginning of this study, I was inspired by Suchman's article [137] on an approach to technology design. I was looking for a way to study the role of CMC in its culture of use with an eye to the implications for the design of such systems. She suggested a transformation of technology design informed by:

1. Recognizing the various forms of visible and invisible work that make up the production/use of technical systems, locating ourselves within that extended web of connections, and taking responsibility for our participation;
2. Understanding technology use as the recontextualization of technologies designed at greater or lesser distances in some local site of practice;
3. Acknowledging and accepting the limited power of any actors or artifacts to control technology production/use;
4. Establishing new bases for technology integration, not in universal languages, but in partial translations;
5. Valuing heterogeneity and partial integration, achieved through practices of technology production/use, over homogeneity and domination. [137]

The notions that there were visible as well as invisible forms of work, that the use of the technology could in any way change the technology, that this use could not be controlled and that these multiple uses should be valued were captivating. I wanted a methodology that would encourage this view of Usenet use. Use was no longer a static dimension but a dynamic process which transforms the design to fit the needs of the people who use it.

In reading Akrich [4] and Latour [75], my approach expanded again. Akrich argues that technical objects are a part of and participate in the building of heterogeneous networks among people, products, and tools that bring these actants together [4]. An actant is defined as "whatever acts or shifts actions, action itself being defined by a list of performances through trials; from these performances are deduced a set of competences with which the actant is endowed" [5]. This view of technology requires a constant movement between the technical and the social.

If we do this, two vital questions start to come into focus. The first has to do with the extent to which the composition of a technical object constrains actants in the way they relate both to the object and to one another. The second concerns the character of these actants and their links, the extent to which they are able to reshape the object, and the various ways in which the object may be used. Once considered in this way, the boundary between the inside and the outside of an object comes to be seen as a consequence of such interaction rather than something that determines it. [4]

Thus the computer as a technical object is part of a network of actants. The actants include the people using the computer, the various pieces of software and the hardware itself. Considered in this light, interactions with the computer and on the computer are no longer pre-determined; each subtly shapes the actions of the other.

It was no longer enough to know the purpose of Usenet or how to post an article; I had to be able to understand the various situations in which the system was used--the local and the global senses of its use. It was not enough to look at those things which are easily measured (e.g., frequency of posting) or to look at the context of an activity (e.g., recreation). I had to include the social construction of knowledge within the newsgroup and a mosaic of use. Usenet (and the computer) became a malleable entity as did the uses, users and situations within any particular instantiation of use.

These ideas coalesced into a study to discover a grounded theory that explains how computers are used as tools for connectivity and community and how one newsgroup had formed a community. Instead of testing a theory or hypothesis, I apply a qualitative approach to make sense of CMC use in an electronic group setting, asking how and why things fit together. Because my study is influenced by Baym's practice-oriented study of a newsgroup [12], I begin this chapter with a description of practice theory. In addition, I include a description of activity theory. Although compatible, these approaches place emphasis in slightly different arenas. Together they provide insights into how and why AGM has built a community. In the remainder of the chapter, I describe qualitative research methodology and my approach to the study of AGM.

Practice Theory. Practice is defined as "participation in an activity system about which participants share understandings concerning what they are doing and what that means in their lives and for their communities" [77]. The activity system is composed of an experienced world (i.e., persons-acting, setting and activity) and its constitutive order (i.e., the relation between systems of belief and social institutions) and the relation between them [76] as illustrated in Figure 1 [76]. A community of practice is then a set of activity systems over time and in relation with other activity systems [77]. Because culture is defined in terms of shared institutions, understandings, membership and actions [12], rather than physical location [77], practice theory allows consideration of cultural processes in non-traditional groups. AGM is a group of people using a newsgroup to communicate. Thus, if AGM develops a shared understanding of their activity and what that activity means, it forms a community of practice.

Systems                               Social
of Belief -----------------------  Institutions
                    |
                    |                         Constitutive Order
- - - - - - - - - - | - - - - - - - - - - - - - - - - - - - - - -
                    |                         Experienced World
                    |
       Persons-     |
        acting ----------- Settings
              \           /
               \         /
                \       /
                Activity

Figure 1. Activity System

Even though there is no one practice approach, Baym asserts there are "two aspects of the common ground of practice theory: the agreement that the generative mechanisms of culture are located in communicative practice, and, correspondingly, the attention to how social meanings are embedded and created in practice" [12]. Language is paramount because of its "culture-instantiating force" [12]; in order to be shared, knowledge and understandings require communication. The three key elements of practice theory's conception of culture are recurrent discursive practice, institutions and members of that culture [12]. A community's discourse is a "set of situated, socially-conducted," routine and systematic narrative practices as well as "varied and improvisatory" practices [12]. All practice is situated, occurring in a setting. Practice is therefore structured by the setting and its constitutive order, that is, the systems and institutions in which it is embedded. At the same time, practice influences that order. The members of a community of practice include novices as well as those who are established within the community.

Communities of practice devote themselves to "the generative process of producing their own future" [77]. This process of community reproduction insures that others will perform the activities and carry on the practices of the community. It is "a historically constructed, ongoing, conflicting, synergistic structuring of activity and relations among practitioners" [77]. Novices are then essential to the social reproduction of the community. First, they eventually become the old-timers. Because of them, established members of the community exhibit and explain the systems of meaning that organize the community's culture. In addition, as novices learn to become fully active in the community, they in turn modify the contexts which shape them. "Learning, transformation, and change are always implicated in one another, and the status quo needs as much explanation as change" [77]. Lave and Wenger use the term "legitimate peripheral participation" to explain the role of novices in shaping the culture as they are socialized into it by moving from the periphery of activity toward full participation in practice.

In addition to the generative mechanisms of culture in communicative practice, the technology of practice (i.e., the artifacts used in practice) gives insights into the social meanings and knowledge of the community of practice [77]. More than tool use, understanding the technology of practice provides a connection with the history of the practice. Lave and Wenger further assert that "knowledge within a community of practice and ways of perceiving and manipulating objects characteristic of community practices are encoded in artifacts" [77]. They call the interaction between the technology and the social world in which that technology is embedded a field of transparency. Transparency is not a feature of the technology but is a process involving participation in which the technology performs a mediating function. Transparency refers to the way that using technology and understanding its significance interact to become one learning process. The complex relation between using and understanding artifacts combines the two characteristics of invisibility and visibility in an intricate interplay--"invisibility in the form of unproblematic interpretation and integration into activity, and visibility in the form of extended access to information" [77]. This interplay is explained in their analogy to a window.

A window's invisibility is what makes it a window, that is, an object through which the world outside becomes visible. The very fact, however, that so many things can be seen through it makes the window itself highly visible, that is, very salient in a room, when compared to, say, a solid wall. Invisibility of mediating technologies is necessary for allowing focus on, and thus supporting visibility of, the subject matter. Conversely, visibility of the significance of the technology is necessary for allowing its unproblematic--invisible--use. ([77], p. 103)

Practice theory, then, supplies a basis for studying CMGs as communities of practice. Baym claims that it is "through the ability of language to index and create cultural meanings that r.a.t.s [rec.arts.tv.soaps, a Usenet newsgroup] and other computer-mediated groups are able to sustain community in the absence of spatio-temporal commonality" [12]. The study of a particular CMG, its discourse and the relation of the discourse to the group's cultural purposes provides a clearer conception of the motivational side of practice. Cultural knowledge and cognition are embedded within the communicative practices of the community and these practices form the foundation of the community. Special attention also needs to be paid to the technology of practice. In a CMG, computer use is the primary method for participating in that community and thus the computer becomes as invisible as a window. The transparency of computer use in CMC needs to be made explicit and visible in order to understand the relationships among the knowledge, the history and the cultural life of the community of practice. Chapters III through VII explain the role of the computer in CMC in the context of AGM, removing its transparency and making it explicit and visible.

Activity Theory. Activity theory refers to the Soviet cultural-historical research tradition in psychology begun in the 1920's [73]. It focuses on practices as development processes. The objective is to understand the unity of consciousness and activity, incorporating "strong notions of intentionality, history, mediation, collaboration and development in constructing consciousness," which is "located in everyday practice: you are what you do" [97]. An activity contains relationships between a subject, object and community. "An activity is a form of doing directed to an object" [73]. At an individual level, an activity is composed of a relationship among a subject, an objective, and tools which is transformed into an outcome. The subject is a person or a group doing the activity; the object or objective motivates the activity; the tools mediate the activity. These ideas can be expanded to supraindividual units [69] by adding community to the relationships [73]. Community refers to those who share the same object. Thus, there are three mutual relationships among subject, object and community which are mediated by tools, rules and division of labor which are engaged in the transformation process of the object into the outcome. These relations are diagrammed in Figure 2 [73]. A tool is anything used in the transformation process including instruments, signs, and language; rules are explicit and implicit norms, conventions and social relations within a community; division of labor refers to the explicit and implicit organization of a community, the roles and communication procedures within the community [15]. These mediating artifacts are created by people and carry with them a culture and history [97]. An activity is a systemic whole--all elements have a relationship to the other elements [73]. These relationships are bidirectional, e.g., the subject is transforming the object while the properties of the object work to transform the subject. Thus, activities are dynamic and change as the conditions of the relationships change [73].

              Tool
             /    \
            /      \
           /        \
          /          \        Transformation
     Subject -----  Object -------------------- Outcome
        /\            /\         process
       /  \          /  \
      /    \        /    \
     /      \      /      \
Rules----- Community ----- Division of labor

Figure 2.  Basic Structure of an Activity.

An activity is a long-term transformation requiring several steps or phases [73]. The short-term processes consist of actions which have operations. Actions are goal-directed processes that are conscious efforts because the goal must be kept in mind. On the other hand, operations can become routinized and unconscious with practice. As conditions change, operations may revert to the level of conscious action until they become well-defined habitual routines. "This action-operation dynamics and the broadening scope of actions is a fundamentally typical feature of human development" [73].

While the historical development of activity and the mediating role of artifacts is important in activity theory, only those aspects of culture, values, motivation, emotion and human personality that are related to the rational understanding of human interaction with the world are considered [69]. Even so, persistent structures such as institutions and cultural values shape activity. Activity theory is not meant to be used to define and understand culture, but is suited as a starting point in studying contextually embedded interactions [73].

Yet activity theory is distinct from situated action models of interaction. Situated action models concentrate their attention on the action and operation levels, on the way people orient to changing conditions but without the idea of an object's ability to influence and generate activity [97]. Suchman calls goals (objects) and plans "retrospective reconstructions" [136] and stresses the primacy of the situation. The subject's reaction to and orientation in a situation are what determine action. Activity theory goes beyond the particularities of any immediate situation to include the "consciousness, intentionality, plans, motives, and prior knowledge as critical components of human thought and behavior" [98]. Context then has a special place in activity theory:

Activity theory, then proposes a very specific notion of context: the activity itself is the context. What takes place in an activity system composed of object, actions, and operations, is the context. Context is constituted through the enactment of an activity involving people and artifacts. Context is not an outer container or shell inside of which people behave in certain ways. People consciously and deliberately generate contexts (activities) in part through their own objects: hence context is not just "out there." [98]

Usenet interactions are contextually embedded interactions. A computer is a tool mediating human interaction with the world. This indicates two relationships or borders, that of the user from the computer and that of the user and computer from the outside world including other objects and other people. Activity theory terms the functional, goal-oriented integration of internal (mental processes) and external (visible) resources as a functional organ. Therefore, a user and a computer can form a functional organ when the external tool is experienced as a property of the individual [69]. This has immediate implications in the study of human-computer interaction (HCI) and computer-supported collaborative work (CSCW). Tools, the culturally developed ways of using them, the mediation which shapes activities and the internalized use of those tools influence the nature of internal activity, shape the goals of the people who use them and through mediation transmit cultural knowledge [69]. Explicit as well as implicit objects or goals are built into tools and tools mutually influence the people using them and the objects of the activity. The explicit object of Usenet was a forum for discussion of Unix [114]; the implicit goal was communication. From these goals, people have used the Usenet as a tool in developing other goals, such as a rebirth of community [114], research collaborations [138] and establishment of interpersonal relationships [12].

Nardi asserts that "a creative synthesis of activity theory as a backbone for analysis, leavened by the focus on representations of distributed cognition, and the commitment to grappling with the perplexing flux of everyday activity of the situated action perspective, would seem a likely path to success" [98] in HCI studies. Since HCI is a component of CMC, an application of such a synthesis to CMC research would also "seem a likely path to success." Baym also argues for a more comprehensive approach in practice theory research. Part of her study was to explain the purposes of practice in r.a.t.s, an analysis of which would "inform practice theory insofar as it elaborates the relations between goals, affects, motivations, discourse and consciousness" [12]. Her research aims to "clarify one of the murkier aspects of practice theory" by elaborating on "the ways participants position themselves in terms of the social world when shaping their own identities, the roles participants can take, the roles in which they place others, and the rights, obligations and power dynamics incumbent in these roles" and "the ways participants create solidarity and affiliation as well as asymmetries" [12]. Proponents of each approach, Nardi for activity theory and Baym for practice theory, affirm the need for a more comprehensive approach to CMC research.

Neither practice theory nor activity theory provide a set of procedures or techniques with which to conduct research, yet both rely on ethnographic methods. Nardi lays out four practical methodological considerations for HCI research. First, the research time frame should be long enough to understand the users' objects including changes in those objects over time and the relationships between those objects and the objects of others in the environment studied. Secondly, attention should be given to broad patterns of activity which disclose the overall direction of an activity and not by looking at episodic fragments in isolation. Third, a variety of data collection techniques should be used including interviews, observations, video and historical materials. Finally, the research should be committed to understanding things from users' points of view [98]. These considerations are the same as laid out by Baym in her practice approach to r.a.t.s--that the research should be "flexible and self corrective, sustained and engaged, and simultaneously microscopic and holistic" [12].

My background reading did not start with practice theory or activity theory. It began with readings in ethnomethodology by Garfinkel [45] and in the sociology of technology by Bijker and Law [17] and MacKenzie and Wajcman [87]. Although I understood the mechanics of Usenet from a computer science perspective and from a user's perspective, I did not know how to study the use of Usenet. I had an intuitive sense of the interrelatedness of the technology and the social system within which it existed. Reading that "the compelling nature of much technological change is best explained by seeing technology not as outside of society . . . but as inextricably part of society" [86] reaffirmed my belief in these interrelationships. Even more so, Akrich argues that the technical content of a technology inscribes a definition of projected users and their use of the technology which must be continually compared with the real user and their uses of the technology [4]. New relationships among Usenet, the participants and the computer became visible and exciting. To understand the effects of Usenet or any technology on society would "require an understanding of the overall dynamics of a society" [86]. Therefore, Garfinkel's ethnomethodology, or the study of practical actions [45], seemed a good starting point.

This next section of the chapter is a review of qualitative research and methods that framed my study. Qualitative research includes categories of ethnomethodology, ethnography of communication and holistic ethnography [88]. I discuss each of these research approaches in turn and then detail the qualitative method I used in my study of AGM. Garfinkel established ethnomethodology as a "method for discerning the taken-for-granted background assumptions, tacit knowledge, behavioral norms, and standard expectancies through which participants constitute social scenes and routine interactions" [85]. He uses the term "to refer to the investigation of the rational properties of indexical expressions and other practical actions as contingent ongoing accomplishments of organized artful practices of everyday life" [45]. Thus, in the study of how people organize their everyday lives you can detect the ways they create social organization [82] and universal ways of making sense [105].

Because conversation is a social achievement [82], conversation or discourse analysis is a related development in ethnomethodological research [85]. Issues in conversation analysis include the features of conversation, the rules and the coherence of conversation. Features are such things as turn-taking and silences. One assumption about conversations is that they are stable with orderly turn-taking [82]. Silence is a communicative activity which is often associated with sociopolitical oppression and/or control [65]. The study of conversational rules include guidelines for standard types of conversation, violations of rules and prevention and repair of errors in talk. Grice has proposed a general rule of conversation which he calls the Cooperative Principle--"make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose of direction of the talk exchange in which you are engaged" [47]; in other words, your contribution must be appropriate [82]. Four general categories or maxims are included with the Cooperative Principle: Quantity, Quality, Relation and Manner. The Quantity maxim states that your contribution should be as informative as needed but not more so than required; the Quality maxim, your contribution should be truthful; the Relation maxim, your comment must be relevant; and the Manner maxim, your comments should be lucidly presented, neither obscure, ambiguous nor disorganized [47]. Conversational coherence theories deal with how conversations are made well-structured and sensible to participants as a matter of consistency of meaning, an organization of a series of acts or the overall outcome [82].

To study the communication patterns of a group, Philipsen suggests using ethnographic methods. This approach provides a descriptive report of what characterizes a group, "getting to the heart of a peoples' meanings, premises, and rules" [105]. Three major questions emerge from this type of study--questions of norms, forms and cultural codes. Norms refers "to the ways communication is used to establish a set of standards and to the ways notions of right and wrong affect communication patterns; forms, "to the types of communication used within the society;" and cultural codes, "to the meanings of the symbols and behaviors used as communication in the cultural community" [82].

Ethnography is also suggested when designing computer artifacts intended to support the activities of the described communities [18]. In linking the ethnographic approach with design activities, Blomberg et al. begin with a description of four main principles that guide ethnographic work--natural settings, holism, descriptive and members' perspective. The research is conducted in the everyday environment (natural settings) where people carry out their activities rather than laboratory or experimental settings, in the belief that those particular activities can be understood only by describing the context which gives the details their meaning (holism), describing "how people actually behave, not how they ought to behave" (descriptive) and how they "organize their behavior and make sense of the world around them" (member perspective) [18].

How CMC is being used by AGM to build community is my central question. Garfinkel's work suggested that an answer could be found by studying "the taken-for-granted background assumptions, tacit knowledge, behavioral norms, and standard expectancies" [85] of AGM. Included in the background assumptions of AGM are computer literacy, knowledge of Usenet structure and norms, as well as AGM-specific expectations. Because discourse is the basis of interaction on Usenet, an ethnography of the communication patterns of AGM communication has a prominent place in my study. Lastly, a holistic, descriptive study in the natural setting, which emphasizes participants' points-of-view, offers a guide to an answer.

The processes of engagement in a text-only, virtual community are little understood; therefore, a methodology is necessary that allows the exploration of the situations of use and that has the flexibility to handle unforeseen processes and interactions. Qualitative methods provide this flexibility [88]. Qualitative research strategies use traditional data collection techniques of participant observation, interviews and recording of field notes. Data for qualitative research typically comes from fieldwork where a researcher spends time in the setting, observing, taking field notes, interviewing and possibly video-taping the participants at the site in which the activity occurs [38]. Although Usenet has no place in physical reality, a newsgroup provides a shared environment of complex interactions mediated by a computer. Qualitative methods can identify and describe the complexity of these social interactions [88].

Participant Observer. A researcher may limit her role in the community under study to that of observer or she may be involved in the daily activities of the community. As both participant in the activities and observer of those activities, the researcher assumes the role of participant observer. The benefits of being a participant observer include firsthand experience with the community practices (which may be the only way to gain access to those activities) as well as gaining a view of the activities from the perspective of a participant [18]. Participation in the activities can help in building trust with the community, as well as provide insights to community norms and relations and an in-depth view of community life [88, 141].

Being a participant observer also has liabilities. One problematic occurrence is that of "going native" [18] which implies a loss of objectivity. Additional problems in on-line activity are undisciplined observations and limiting observations to those in which the researcher was directly involved [12]. Hiltz recommends summarizing the objective data and minimizing any acquired biases as the main solution to going native [58]. In his study of prisons, Thomas assesses his involvement with the prisoners through reflection.

Reflection refers to the act of rigorously examining how this involvement with our subjects affects our data gathering, analysis, and subsequent display of the data to an audience. Through reflection, an act of repeated thinking about our project, we attempt to become self-aware of the process and consequences of knowledge production by bringing the original act of knowledge back into consciousness. ([141], p. 46)

His solution was to write about topics that were less likely to be affected by his empathy for the prisoners who had become his friends. Instead of attempting to maintain objectivity, another approach is to explain the positionality of the researcher. Wolf states that "positionality is not fixed, but relational, a constantly moving context that constitutes our reality and the place from which values are interpreted and constructed" [153]. She puts forth a research perspective that not only encourages researchers "to bring their own particular location and position into the research, to acknowledge and build on their partial perspective, but makes it imperative for them to do so before any discussion of another's reality can be introduced" [153]. Lave echoes this stance:

I seek to explicate the manner in which my positionalities within those locations is implicated in the account that I provide or, in other words, that 'the problem of voice (speaking for and speaking to) intersects with the problem of place (speaking from and speaking of)'. [76]

Because I have been a part of AGM as a lurker, infrequent poster, attendee at AZAGM95 (the first annual international meeting of AGM held in Prescott, Arizona in August, 1995), I have the position of participant observer in this research. I own several AGM t-shirts and regularly exchange postcards, email, letters and audiotapes with some of the AGM participants. I know some of the history of AGM, I recognize most of the people who post by their userid and I am recognized as a member of AGM. The people who regularly participate on AGM are my friends. I have received emotional support from the group and I met my husband on the group. Therefore, my involvement in this study is significant. My bias in this study has been that I believe AGM is an active and viable community.

As participant observer, I have the opportunity to influence the discussion on AGM. Yet, I consider myself more of a lurker than a regular contributor to AGM because I post only occasionally. By continuing my usual posting habits, I believe that the practice and discourse on the newsgroup has remained in its natural state even through the data collection phase of the study. I have shared my work with participants who are interested but I have not raised its contents as a topic of discussion.

There have been benefits to my being a participant observer. Learning to distinguish among the individuals who regularly post to AGM, learning who is new (a newbie) and who is merely posting after an absence (de-lurker or oldbie), as well as who is an interloper can take time and effort. Because of my long-term participation in AGM, I had already invested time in learning the skills necessary for Usenet participation together with the identities of AGMers. Building trust is another time-consuming effort. Again, as an established member of AGM with close friends among the group, I had only to maintain that trust. Toward that end, I posted several announcements of the study, explaining my reasons for conducting the study and the uses that would be made of the results. Although posts to Usenet are considered public, I obtained permission from the individual posters when including their posts in the dissertation, as well as actively involving them in the choice of pseudonym and description for their persona. The trust invested in me by AGM is evident in the large response I had to the questionnaire. Of the approximately 200 AGM posters, I received 112 completed forms. My role in AGM as researcher and observer, albeit similar to lurker, has been recognized for its objective perspective; I was asked to respond to a concern among some AGMers about the effects of an influx of newbies if a news story were to be printed about AGM. Because rereading Usenet discourse is a different experience than reading the posts as they are received, an additional benefit to being participant observer is my ability to recontextualize the discourse during the analysis phase of the study.

Informational Adequacy. Even though I began this study with the belief that AGM is an active and viable community, I have not limited my data collection and analysis only to those interactions that would support that belief. I collected a wide range of detailed data, including the daily posts, over a three-month period from October 21, 1995 through January 31, 1996. This length was necessary to provided analysis over peak as well as quiet periods and tumultuous as well as peaceful periods that typify newsgroup activity. During this time I received 9,839 posts to AGM. One of the advantages of studying CMGs is that the public interaction need not be transcribed from field notes; it is already recorded and only needs saving. Other data I collected include interviews, email interaction, traffic and readership statistics and an electronically distributed questionnaire.

My interviews with AGM participants were conducted primarily via email although one interview was conducted face to face (FtF) when a European AGMer visited in Texas. I used an electronically distributed questionnaire, transmitted via a posting to the newsgroup and email, to gather demographic data and personal statements about the purpose of AGM, favorite AGM stories and why the participant joined AGM. This questionnaire was distributed after the three-month posting collection period to minimize any influence on the discursive practices of AGM. I also gathered additional statistical data to determine basic characteristics of the environment, such as number of daily messages, posting patterns of participants and lengths of threads. By collecting detailed and repeated samples from multiple data sources, I have done my best to continue my involvement with AGM and still remain reflective and flexible in my analysis.

Ethical Considerations. Privacy and ethical considerations are of primary importance in this type of research [88]. Usenet research includes the usual ethical concerns of informed consent, freedom from harm, privacy, and confidentiality as well as concerns of copyright infringement. However, it is difficult to apply these principles to the study of Usenet postings. Informed consent is usually obtained before a person participates in the research. It is not possible to contact and obtain such permissions before the person posts a message. Even posting a notice of the research does not guarantee that all posters will receive the notice or will read the notice. A Usenet newsgroup is more than the active participants who post; there are hundreds of lurkers [110]. AGM has an estimated potential readership of 6,100 people [7] yet the number of active participants ranges between 200 and 300 in any month. Usenet is a public forum similar to any busy street corner; posts are considered public domain [70]. Privacy in a newsgroup is guaranteed only to lurkers. Whether Usenet postings are protected by copyright has not been settled. That some of the posters use pseudonyms and that some of the postings are of a personal nature are evidence that privacy and confidentiality are necessary.

Because of the distributed nature of the newsgroup in both time and space, I posted multiple announcements to inform the AGM community of the study. Strict confidentiality has been maintained throughout the study; the computer files where the data is kept has permissions set to exclude everyone except myself. In this dissertation, the identity of participants (i.e., name, email addresses and affiliations) is masked by pseudonyms. Following the lead of Turkle [145], I have asked AGM participants whose work I have quoted how they wish to be identified, i.e., their pseudonyms and descriptions. (One poster who normally uses a pseudonym found giving his pseudonym a faked identity hilarious.) A summary of the survey data has been posted to the newsgroup for public discussion and is available on-line from my Web page. As this dissertation has been written, it has been made available electronically to AGM by a variety of methods including posting to the newsgroup, emailing to those requesting it and providing links from my Web page. By making the dissertation public to AGM, I have provided them the opportunity to verify any statements made, to correct any mistakes and to reassure themselves that I am sensitive to any of their concerns for privacy and confidentiality.

My motivation for this study has influenced the selection of a theoretical and methodological framework from which to work. Discourse is the basis of interaction on Usenet; therefore, discourse analysis has special importance in the study of a newsgroup. Although Usenet conversation does not have turn-taking as usually experienced in a FtF conversation because of the spatial and temporal differences among the participants, other features of conversation are apparent. Silence is perceived and used in AGM (covered in the chapter on discursive practices). The Cooperative Principle with its four maxims is also much in evidence. Just as violation of any of the maxims causes a disruption in the flow of a spoken conversation or affects the perceptions of others in the conversation [82], violations of the maxims affect the communication on Usenet. A violation of the Quality maxim caused a major disruption in AGM interaction (discussed in the chapter on community). Conversational coherence has significance in Usenet. When discourse and meaning seem coherent and you perceive yourself as "coherent in interaction constituted by the discourse, [it] creates an emotional experience of insight (understanding the text) and connectedness (to other participants, to the language, to the world)" as well as sending "a metamessage of rapport between the communicators" [138]. Conversational coherence is therefore necessary for personal involvement in a newsgroup and the sustenance and maintenance of the community in a newsgroup.

My purposes in this study are two-fold, descriptive and explanatory. I have used participant observation, interviewing, a questionnaire and document analysis as the basis of those descriptions and explanations. The remainder of this dissertation is a description of the contexts in which the community formed by AGM exists and an explanation of the forces that have contributed to the formation and continuation of that community