For this project, I used qualitative research methods common to ethnographic research and additional historical library research, as well as methods particular to the electronic medium on which my subject of research resides. James Spradley defines ethnography as "the work of describing a culture. The central aim of ethnography is to understand another way of life from the native point of view." (Spradley, 1980: 3). In general, ethnography allows us to examine cultures different from our own, and to learn from them things that might not be apparent to the casual observer.
Participant observation is an important technique central to ethnographic research. As a participant observer, the ethnographer is able to gain the perspective and understanding that comes from being an "insider" to the culture. While participating in the activities of the culture, it is also imperative for the ethnographer to maintain a critical and "observational" attitude. She must attempt to experience the culture as both insider and outsider and to allow her own observations and experiences to also be informed by the other participants in that culture. In the field of anthropology, ethnographic fieldwork was traditionally conducted in a culture very different from the one to which the ethnographer was native. Often, research was done in an attempt to learn from and better understand traditional cultures and societies that were rapidly changing, disappearing, or being absorbed into larger, mainstream cultures surrounding them. More recently, however, many ethnographers have been doing research within their own society. This often looks at subcultures inside one's own culture, such as Michael Agar's study of New York City heroin addicts (Agar, 1980), or Oliver Sacks' research on the Deaf and their unique culture (Sacks, 1990). Paul Willis' (1977) account of British working class students gives us an example not only of looking at subcultures, but of actually studying one's own culture.
Ethnographic research in cyberspace presents a number of challenges; since the members of one virtual community may, in "real life" be scattered around the globe, it is very difficult for the ethnographer to meet with or interview all of them in person. For this reason, interviews and other data may be collected "virtually". However, this method brings up another set of potential problems. In cyberspace, an informant can easily disguise many things about herself - gender, age, ethnicity, and most of the other variables normally considered important by anthropologists. For these reasons, I explored and present here some varying opinions on doing ethnography of cyberspace before I go on to discuss my own methods.
Dr. Sherry Turkle (1995), in researching her own work, spent a great deal of time observing, in person, the way that people interact with computers. She worked in places like computer labs, classrooms, and scientific conferences. Turkle has chosen not to present her findings unless she has met the informant in person. Her reasons for this include:
I made this decision because of the focus of my research: how experiences in virtual reality affect real life, and more generally, on the relationship between the virtual and the real. (Turkle, 1995: 324)
However, she goes on to recognize that researchers with different aims may legitimately employ different methodologies:
In this way, my work on cyberspace to this point is conservative because of its distinctly real-life bias. Researchers with different interests and theoretical perspectives will surely think about this differently. (ibid: 324)
Anthropologist Steve Mizrach (1994), in his paper on Computer "hackers" and their language, presents an entirely different point of view. Because his research involved an underground group who were engaged in some illegal or quasi-illegal activities, he found it necessary to conduct his research entirely in the virtual. Beyond this, however, he suggests other reasons for doing so:
From an emic perspective, many of our subjects do not distinguish between "real" life and "virtual" life. As good ethnographers and participant observers, we should not make such seemingly "etic" distinctions in the face of our informants. If they spend more of their waking time in cyberspace than in "real life", who is doing the more honest ethnography? The cyber-ethnographer, or the person who ignores that part of their life to which they devote the most time? ... As good cyber-ethnographers, we should be just as willing to examine the sociocultural relations in "cyber" society as we do in "real" society. ... Where people invest meaning, the anthropological interpreter should go; and people do invest great meaning into cyberspace. (Mizrach, 1994: 1-2)
My research methods for this project were a combination of in-person and virtual. As a participant in these virtual communities, I established characters, interacted with other members and recorded events and conversations where appropriate. However, as an observer and ethnographer, the bulk of my data has come from in-person and e-mailed interviews with members of these virtual communities and with people who are involved with working on them in some capacity.
The purpose of this qualitative study is to provide greater insight into the ways that participants in virtual realities view the cultures and communities they are creating and building. Spradley (1980) advocates the use of participant observation and in-depth interviews to learn about the explicit and tacit knowledge held by individuals while minimizing the risk of imposing predetermined ideas on the population under study. In addition, Kirk and Miller (1986) contend that qualitative research is especially useful when seeking to understand questions of process rather than quantity.
Along with participant-observation, I used the Critical Incident Technique for interviewing (Flanagan, 1954). This is a method of gathering information about important events, or incidents, from members of the virtual communities themselves, in order to better understand how they view the situations at hand. John Flanagan describes this technique as "a set of procedures for collecting direct observations of human behavior in such a way as to facilitate their potential usefulness in solving practical problems and developing broad psychological principles." (ibid: 327). It is a flexible technique, in that it consists of a set of general principles that can be adapted to each situation as needed. There are five steps most commonly used in the Critical Incident Technique (ibid: 354-355). They are:
These steps, and how they have been carried out for this project, are as follows:
The aim of this research project is to learn about formation and enforcement of social and cultural norms in graphical and three-dimensional virtual realities, and to see how this correlates with modern ideas about life on the American frontier of the nineteenth century. By so doing we gain a truer understanding of how the concept of "frontier" is being used to grasp and make sense of electronic medium.
Plans for data gathering were completed during the writing of my thesis proposal. In that phase of the project I primarily relied upon James Spradley's book Participant Observation (1980) along with the Critical Incident Technique documented here for guidelines. As Spradley says, "In ethnographic inquiry, analysis is a process of question-discovery. Instead of coming into the field with specific questions, the ethnographer analyzes the field data compiled from participant observation to discover questions. You need to analyze your fieldnotes after each period of fieldwork in order to know what to look for during your next period of participant observation." (Spradley, 1980: 33-34).
Therefore, my plan was to develop specific interview questions for each virtual world I explored, as I went. I began by establishing characters in each world and logging in for a time to begin to learn about the general character of each one. I also looked up background information about each one, using published histories and other papers, mailing lists, and online asynchronic discussion groups. The information gathered then assisted me in generating a set of basic questions relevant to each realm.
For the purposes of this project, interview questions centered on incidents that have occurred within virtual communities and the development or enforcement of social norms therein. With the Critical Incident technique, it is important that the questions be as specific as possible. Additional questions arose as I interviewed my informants and acted as a participant observer in the worlds. See Appendix D for lists of the basic questions I used pertaining to each world.
In total, I conducted thirteen interviews: one on the phone, six in person, and six through e-mail. Each interview began with the specific questions that I had in mind, and many of them moved into additional discussions of topics and issues that were introduced by my interviewees. Overall I found all of these interviews very useful and informative. I did, however, notice some big differences between interviews conducted live and those conducted through e-mail. During the live interviews there was far more tendency for the subject and conversation to wander into other topics, as each informant discussed those aspects of virtual worlds and cyberspace that he or she found most engaging. As a result, my taped interviews were very lengthy and very interesting, but I had to do more digging to get to the specific kinds of comments and data that I was looking for. By contrast, my e-mailed interviews remained much more focused on the questions I originally asked, and on the discussions arising from them specifically.
Data was also collected in a number of other ways, from a variety of sources. One of my first activities as I began my research was attending a conference held in October of 1996. "Earth to Avatars" was held in downtown San Francisco, and was presented by a group called The Contact Consortium . This organization is dedicated to the study of virtual worlds, and its membership is multidisciplinary, consisting of engineers, social scientists, writers, musicians, graphic artists, universities, corporations, and many other interested parties. Earlier in the year I had attended an event they sponsored in AlphaWorld at a special site called "Sherwood Forest"; the purpose of this event was to begin to collectively build an environment, and to discuss communication and community in avatar space .
At the Earth to Avatars conference I attended a number of panels, including one on virtual communities. It was led by Wendy Sue Noah, the founder and chairperson of WEAVE , the women's special interest group of the Contact Consortium. There I met Wendy Sue and asked to become part of WEAVE. Through my involvement with WEAVE, I attended the first organizational meeting and became WEAVE's Web site developer. I also attended another event consisting of presentations and talks on virtual worlds, including the three worlds I have explored for this project. Through this organization I met and spoke with many women who are interested in social development in virtual worlds, several of whom, including Wendy Sue, became informants.
In each of the worlds I studied, I created a character and went in to learn a little about the world - how it operated, what the general "mood" was like, how people interacted and related to each other and to their environment. My avatars, each known as "Christy", were identified as female characters wearing red. The appearance of one's avatar can have a profound effect on how you are perceived and treated in cyberspace, and my decision to enter these virtual worlds as female, humanoid characters was primarily based on my desire to be treated in virtual space as closely as possible to how I would be treated in real life.
In addition, I began to participate in online forums, news groups and mailing lists pertaining to each world. These were a great source of information, as people were in there discussing many of the social and behavioral issues that I was attempting to learn from. Although I did not conduct research directly in these forums, I did find the majority of my remote interviewees through them.
The discovery and analysis of cultural themes as they exist in virtual worlds and correlate to those found on the American Frontier form the core of my data analysis. Anthropologist Morris Opler defines a "cultural theme" as "A postulate or position, declared or implied, and usually controlling behavior or stimulating activity, which is tacitly approved or openly promoted in a society" (Spradley, 1980: 140).
My analysis followed James Spradley's methods (ibid) of domain analysis and identification of dimensions of contrast, and using that information, seeking out cultural themes as they began to emerge. In Chapter 5 I will go through specifically how I used these methods.
For this step in the Critical Incident Technique the researcher must attempt to report any possible biases and limitations that have arisen during the previous four steps. In addition, the degree of credibility of the final results should be pointed out.
Firstly, as with any research conducted in virtual space, there is the issue that with my remote interviewees, I had no way to verify that they actually are who they claim to be! Although I asked each of them for some brief demographic information (gender, age, and place of residence), I have only their word to go on. In addition, the informants I have spoken to were selected from among people who had professed an interest in discussion of social issues around cyberspace. I met my informants, both face to face and virtual, through meetings and forums around such issues, and my requests for interviews were generally gladly granted.
Partly to address the issue of credibility of my analysis, I established a web site dedicated to my research . On this site, informants as well as other interested persons could view my work as it progressed, make additional comments, and provide feedback through e-mail. In addition to providing some level of reciprocity to the communities that I was researching, I was hoping that this would assist me in keeping my analysis and conclusions on track.
Although I encouraged all of my informants to review this site and to provide me with any comments or feedback they had, I heard back from only a couple of them, along with a few other interested people who heard of the site and my research from other sources. However, the comments I received were helpful and did, indeed, assist me in ensuring that my thoughts and conclusions were valid. Among these were "Spring Dew", one of my informants who was also the editor of the "New World Times". From reading my posted proposal, she was able to see that my background research had occurred before AlphaWorld was bought by Circle of Fire. This was helpful information, and I was able then to look into the transition of ownership and how it had affected the AlphaWorld residents. This transition is documented further in section 3.2.1, where I discuss the history and background of AlphaWorld.
Also responding was Cathy White, a student at Kent State who found my site while searching the web for information about Frederick Jackson Turner's work. Her own research was also involving the Internet and Turner; as she told me, "My research involves the use of the Internet by the "fringe" elements of society as a frontier in a society that often rejects extremists" (C. White, e-mail communication, 1997). I thought it was interesting that someone else was also looking at the Internet environment as a frontier, and since her research was focused on deviant behavior, I referred her to the works of Suler (1997) and Mizrach (1994) as sources that I had found helpful in that area.
Another person to offer comments was Mark Caban, an Australian member of an e-mail list that my husband Henry belongs to. Henry had at one point made some comments on the list that referred to my thesis, and after looking over my site, Mark felt that "The parallels with the American frontier can be transposed to the colonial society that existed in Australia in the 1880s" (M. Caban, e-mail communication, 1998). One thing I have mentioned in section 6.2 (Opportunities for Further Research) is the need to explore whether or not similar parallels could be seen in other countries, and Mark's comments indicate that this might be the case.