Robin Hamman (http://www.cybersoc.com) is a doctoral student and freelance Internet consultant based at the Hypermedia Research Centre, University of Westminster, London. He is currently researching the offline effects of online communication while continueing to edit his critically aclaimed webzine, "Cybersociology Magazine". His past research on topics such as Cybersex Chat, Digital Third Worlds, and Online Community has made him an internationally recognised researcher of cyberspace. He spends much of his online time co-hosting the Life Online conference at Brainstorms. The full thesis upon which this article is published can be found online at http://www.cybersoc.com/mphil/

Many social scientists have written about the development of online communities. Online communities are communities which are formed in cyberspace and, many times, these online communities consist entirely of people who have never met each other offline. In the fourteen years I have spent online, I have been a member of dozens of online communities of this type and have met hundreds of users who use computer networks primarily to meet other users and to build new online relationships with them. However, the research presented in this article suggests that many AOL users first obtain an AOL account to conduct research and to communicate with people from their pre-existing network communities. This breaking down of the boarders between online and offline communities stands in contrast to most existing notions of online communities and points to wider changes important to our understanding of the term "community".

In this article, I present the findings of research I've recently conducted on the users of the popular online service America Online (AOL). The prime focus of this investigation is to look at the motivations of AOL users, and more specifically, to see if social isolation or loneliness has played any part in their motivation to join and use the computer network. The second focus of this study is to investigate how AOL users are actually choosing to spend their time online. The final focus of this research is to investigate how spending time online effects users as individuals as well as their pre-existing offline friendships, social networks, and communities. The findings of this research suggest that many AOL users are motivated to join the online service to conduct research and to communicate via computer with people they already know offline rather than to meet new people online and to build new friendships with them. My findings also suggest that, when we use CMC to communicate with members of our pre-existing social networks, our time spent online may be beneficial to the solidarity of these groups. This stands in opposition to a number of theorists and researchers who have written that computer mediated communication (CMC) can lead to the disintegration of pre-existing communities, a finding which is most notably present in the recent HomeNet study.

Within the social sciences, there is little agreement upon the definition of the term "community" other than it is almost always used to describe a group of people. Poplin writes, "From its inception as a discipline, sociology has been plagued by inconsistency and ambiguity in some of its basic terminology... the word community falls into this category. As an element in the sociological vocabulary, this term has been used in so many ways that it has been described as an omnibus word." (Poplin, 1979, 3) Even the Oxford Dictionary of Sociology states that "the ambiguities of the term community make any wholly coherent sociological definition of communities, and hence the scope and limits for their empirical study, impossible to achieve." (Marshall, 1994, 75)

The term community is commonly used in a number of non-scientific ways within everyday life. This causes terminological ambiguity and confusion when we then try to define the term within the social sciences. Some examples of the diverse ways in which the term "community" is commonly used outside the social science literature include:

• In advertising: "Use your community based grocery store. We create jobs for local people."
• In the names of buildings: "Independence Village Retirement Community"
• To describe a city or it's inhabitants: "Liverpudlians have a strong sense of pride in their community"
• When talking about minority groups such as the "Black Community" or "Gay Community"
• When describing a feeling of togetherness felt by a group of friends or work colleagues.
• When describing abstract groups who are distributed world-wide. For example, the "Academic Community" is made up of thousands of academics who communicate through newsletters, journals, and other forms of interpersonal contact both directly and indirectly.

While the non-scientific usage of the term community leads to misunderstandings when we attempt to use the term in social scientific texts, another important reason for ambiguity surrounding the definition of the term within the social sciences is that the very social construct which it describes is continually changing and evolving. The term has also become value loaded due to the political nature of it's usage in many texts. These changes within the social construct we call community are described by well known theorists such as Durkheim, Marx, Weber, and others, and are caused primarily by technological, socio-economic, and political changes within wider society. For example, Durkheim writes about the changes which occurred when we moved from early rural agricultural societies, largely based upon kinship bonds and shared geography (mechanical solidarity), to modern urban industrial societies where our communities are based upon common interest as well as shared geography (organic solidarity) but not necessarily upon kinship bonds. For Marx, the bonds found in modern communities are based upon money, the quest for it, and the inequalities caused by it's scarcity. Specialised groups, according to Marx, are formed on the basis of class which in itself is determined largely by money and wealth accumulation. For the New Left of the 1960's and 70's, "community" was a way of organising people who were interested in a specific cause. For example, the San Francisco gay community was formed to support a social movement aimed at gaining and formalising the rights of homosexuals. For the May 1968 revolutionaries in France, community was seen as an alternative to capitalist consumerism (Barbrook, 1995, 91 - 114). For just about every social theorist and activist using the term community, there is a political meaning and agenda tied to it.

The above is by no means an exhaustive listing of the uses of the term community over the past century. However, this limited argument does illustrate that there are many competing conceptualisations of the term depending upon who is using it, when they are doing so, and what their purpose is. The development of the written word, mechanisation, increased stratification, constitution, suburbanisation, and so forth have all led to changes within communities and the ways in which we conceptualise them. The term "community" has dozens if not hundreds of distinct definitions in the social sciences. It has been briefly demonstrated here that some of these definitions are based upon political viewpoints. With so many divergent political viewpoints, consensus upon one sociological definition for the term has been impossible to reach thus far. Because of this ambiguity, Freilich instructs us that, "since a requisite of science is specificity of terminology, we must conclude... that at this time `community' is a non-scientific term unless separately defined in every paper which uses it." (Freilich, 1963, 118.) For this reason, it is important to present a clear sociological definition of the term community here before proceeding.

In what may be the best attempt yet to assess agreement amongst definitions of community, George Hillery, Jr. subjected 94 sociological definitions of the term "community" to qualitative and quantitative analysis. (Hillery, 1955, 111) He was able to identify 16 different defining concepts within this sample. (Ibid. 115) Hillery found only one concept that was common amongst these 94 definitions: they all deal with people. (Ibid. 117) Despite this, there are other areas where the majority of studies analysed by Hillery are in agreement. Hillery states that "of the 94 definitions, 69 are in accord that social interaction, area, and a common tie or ties are commonly found in community life." (Ibid. 118) Poplin notes that in a more recent empirical study of 125 sociological definitions of the term community, the above defining aspects were still present in the majority of definitions despite some minor changes in the usage of the term over the years. (Poplin, 1979, 8)

In the interests of keeping this section brief, and to avoid confusion later, I have used Hillary's analysis of definitions of community to come to a single concise definition rather than to present dozens of different definitions here. The sociological term community should be understood here as meaning (1) a group of people (2) who share social interaction (3) and some common ties between themselves and the other members of the group (4) and who share an area for at least some of the time. I choose to deny the term it's historical politicisation in favour of presenting a clear, and hopefully unbiased, definition here. For the remainder of this article, it can be assumed that when I talk about community in the physical world, I am referring to something with all four aspects seen above and that it is a term stripped of it's political meaning.

Later in this article, we will see that community has not disappeared over the years, as some social theorists have written, but has indeed changed. The work of Barry Wellman [see chapters in this volume] and others demonstrates that community is increasingly becoming privatised. Privatised communities take the shape of networks and network communities are more likely to be based upon individuals rather than in neighbourhoods. This shift towards private network communities, from communities rooted in a specific, confined geographic area, is due to the privatisation of public spaces once important to the development of community. In the absence of public gathering places of the type which often facilitate the development of geographically based communities, the internet becomes a practical, efficient, and valuable tool for interpersonal communication which is important to the continuance of private network communities which are based upon individuals.

As the user base for online services and the Internet increases dramatically in number, it is also changing rapidly in so far as it's demographic makeup. Many of these new users are motivated to come online for different reasons than those who came online before them. Just a few years ago most users of computer networks were academics, hobbyists, "netheads", and technophiles - the people the computer industry calls "early adopters". In contrast, many users today are just ordinary people who aren't particularly interested in the technology but who have a job to do and who choose to use computer networks as a tool for doing it.

While computer networks remain entirely inaccessible to most sections of society, for some groups, and within some communities, access to an online service or the Internet is almost as commonplace as having a telephone. Computer mediated communication (CMC) is increasingly being used between members of such communities.

It has been suggested by some sections of the mass media, and by a small number of academics, that those using computer networks do so at a cost to their offline community life. They suggest that by using computer networks, users turn their backs upon friends, family, and colleagues within their local community in favour of communities found entirely online. Within the sample of AOL users who take part in the study presented here, I have not found this to be true. In fact, my findings suggest that users taking part in this study are motivated to use AOL by the need to do research for academic or business purposes and to communicate with others within their pre-existing offline communities. None of the AOL users in my sample report having any interest, need, or desire to meet new people online, to build new online communities, or to take part in their own social experimentation online. Whilst I agree that there are some who use computer networks such as AOL to meet new people and to build new communities online with those they have never met offline, the findings which are presented here suggest that the majority of respondents to this study use computer networks do so to complete specific research or work related tasks and to communicate. When they're done, they log off.

The Internet and other computer networks are increasingly being used within industry, business, and in our personal lives. Computer networks allow us to more easily do many everyday tasks such as information gathering, researching, and communicating. Newspapers, educators, and computer corporations alike have pointed out that online education is allowing geographically remote students to have access to courses taught by the very best experts in their field. (Hamilton & Miller, 1997; Spender, 1995, p.146; Lotus/Harvard, 1998) In a similar vein, it has also been noted that doctors in the West have helped to solve the medical problems of patients in the developing World through consultations on the Internet. (Hu, 1998)

There is no shortage of academic and theoretical critiques which cast a positive light on the Internet and other computer networks. In fact, most researchers who have undertaken empirical fieldwork in cyberspace note that there are a number of possible benefits of participation on computer networks. For example, they write about the freedom of anonymous users to experiment with their multiple selves (Turkle, 1996) and their gender (Stone, 1995) and to gain political power they would not otherwise have obtained (Schmitz, 1997).

Despite all the positive benefits of using computer networks, there are still those who feel that the use of computer networks is detrimental to both individuals and society. For example, it's hard to miss the frequent articles about the availability online of dangerous information, such as bomb building instructions, or about the use of the internet by child pornographers, Neo-Nazi's and other hate groups, and other social deviants. It is fair to characterise the Internet as a dangerous place, but only if we also admit that the library or the streets in front of our homes are at least equally dangerous, and probably more so. Computer networks do allow people to gain access to potentially dangerous information, such as instructions for making fireworks, but this information is also available in bookstores and libraries. Similarly, the chance of children meeting pedophiles online is no higher, and probably much lower, than it is offline. The point is that children probably shouldn't be left alone while they use AOL or the Internet, just as they shouldn't be left alone to watch late night television or allowed to play outside without supervision. Dangers exist everywhere, but there is nothing about computer networks which makes them inherently more dangerous than anywhere offline.

Within contemporary culture it is also sometimes suggested that users of computer networks are technologically minded nerds, many of which have few or no friends. For example, a popular soft drink brand in the UK recently ran an advertising campaign suggesting that trendy people drink their beverage whilst pathetic nerds wear cheap raincoats of the type frequently seen on trainspotters and use the Internet. Across the Atlantic, a store near Ft. Myers, Florida reports that a T-shirt which reads "what part of http://www.getalife.com don't you understand'" is one of their best selling designs. With such views remaining credible within mainstream culture, it's no surprise that the popular media went into a feeding frenzy when the authors of the recent HomeNet study, conducted at Carnegie Melon University, concluded that internet use can lead to increased, self-reported levels of social isolation and depression. Later in this article we'll take an in-depth look at flaws in the HomeNet study which raise serious questions about it's validity.

Others have noted the existence of negative portrayals of computer network users as socially isolated nerds. Howard Rheingold writes that, in the late 1980's, "the media myth about people who used computers to communicate was that we were pencil-necked nerds, totally lacking in social skills, whose online communications are robotic and unemotional." (Rheingold, 1998) This "myth, as Rheingold calls it, has influenced the work of a number of social theorists and researchers.

Michael Heim warns of the danger that reality will be supplanted by simulations as was recognised by Jim Morrison of The Doors who wrote in 1969, "There may be a time when we'll attend Weather Theatre to recall the sensation of rain." (Heim, 1993, 82) Heim later warns that "Technology increasingly eliminates direct human interdependence. While our devices give us greater personal autonomy, at the same time they disrupt the familiar networks of direct association." (Heim, 1993, 100) Heim fails to prove or demonstrate that there is any validity to this statement.

In Data Trash , Kroker and Weinstein warn of the increased loneliness of the online community member when they write,

While Heim, Kroker and Weinstein write powerful warnings about the dangers of online communication, or as Kroker calls it our "will to virtuality", none of these have put their theories to test using empirical social scientific methodology.

The recent HomeNet study undertaken at Carnegie Melon University in the United States has been widely reported in the media. The results of the study even made the front page of the Sunday New York Times (25 Oct. 1998) with the headline, "Sad, Lonely World Discovered in Cyberspace". Similar headlines were used by all the major news agencies including AP, CNN, and the BBC. According to the Carnegie Melon researchers, the study "examines the social and psychological impact of the Internet on 169 people in 73 households during their first one to two years on-line." Their findings suggest that, "greater use of the Internet was associated with declines in participants' communication with family members in the household, declines in the size of their social circle, and increases in their depression and loneliness. These findings have implications for research, for public policy, and for the design of technology." (Kraut, etal, 1998) Despite these claims and the great media attention given to them, "the statistically significant changes the researchers report are quite small -- like a 1 percent increase on the depression scale for people who spend an hour a week online." (Rosenburg, 1998)

Whilst the authors of this study should be applauded for undertaking such important and timely research, their research design is seriously flawed in a number of ways. These design flaws have caused a number of well known authors to question the validity of the findings of the study.

The HomeNet study is based upon data collected from a sample consisting of 256 members of 93 families drawn from four schools and several neighbourhood groups in the Pittsburgh area. Respondents are members of families with no previous access to computer networks from their homes who were given free computer equipment, a second telephone line, Internet access, and training in exchange for taking part in the study. This free equipment and training, worth several thousand dollars (£1200+), was probably their main motivation for going online and for taking part in the study. In contrast, most users of computer networks have certainly not been motivated to go online by the promise of free computer equipment and training for doing so. Because respondents taking part in the HomeNet study have an entirely different motivation for going online than most users of computer networks, the findings of the study certainly cannot be generalised outside of the population studied. Furthermore, it is possible that some participants in the HomeNet study may have been unduly influenced by the training which they received prior to the study. For example, if they were taught to use the Internet to access anonymous chat rooms at Yahoo or to visit entertainment websites instead of using their access to join in real online communities or to keep in touch with people they already know from offline, it is very possible that this training caused them to use the Internet in ways which led them to feel more socially isolated. The HomeNet study ignores the different motivations for going online which are exhibited by the diverse population of Internet users, and fails to recognise that the gift of several thousand dollars worth of computer equipment, access, and training could have any number of unexpected and unknowable consequences upon the ways in which respondents choose to use the Internet.

Another serious flaw of the HomeNet study is that it's findings are based upon data collected during only a pre-test questionnaire and a single subsequent questionnaire which was administered between the 12th and 24th month of the research. Considering that the HomeNet study is meant to be a longitudinal one, there should have been more questionnaire tests administered to assess the validity of findings. Outside factors which could have caused members of the sample population to report "changes in their social involvement and psychological well-being" between the 12th and 24th weeks of the study were not investigated. In his critique of the HomeNet study for Salon Magazine, Rosenburg writes that "the researchers only tested people twice... which doesn't provide a very wide set of data points to offset the impact of other factors (time of year, state of the economy, random personal crisis)." (Rosenburg, 1998) The researchers responsible for the HomeNet study could not possibly have controlled all outside variables, such as the local economy or weather, and further steps should have been taken to investigate the possibility that such outside factors may have led to the slight increase in respondents self-reported levels of depression and social isolation. In fact, Rosenburg writes, "the researchers have no idea whether their subjects got bummed out because of what they encountered on the Net, or simply because they wound up sitting in front of a computer monitor rather than working in their gardens or playing ball. Is the increase in "loneliness and depression" caused by the Internet itself or simply by computer use, regardless of whether the modem's on' The study can't say." (Rosenburg, 1998). Regardless, a one percent increase in loneliness and depression, as measured through the self-reports of participants, is hardly significant, especially when dealing with such a small sample of users.

Another possibility, not discussed in the study, is that using the Internet may simply have led members of the sample to become aware of already existing deficiencies in their offline relationships. Online communication can often lead to improved self-awareness since people have been known to be very open and self-reflective whilst online. (Hamman, 1996) This new self-awareness could lead to reports of increased levels of depression, isolation, and loneliness among respondents. Similarly, having many friends online, but few offline, could lead to an awareness that something is missing from respondents offline lives. Because the HomeNet study is flawed methodologically, it is impossible to conclude that there is a causal relationship between using the Internet and increased levels of depression and loneliness.

Another serious flaw of the HomeNet study is that it assumes that relationships conducted offline through face-to-face interaction are qualitatively better than relationships conducted online through computer mediated communication. The study uses the existing (pre-test) level of communication within sample families as a benchmark, and assumes that increased levels of communication between these family members is always positive. This benchmark is inappropriate since many people find much to be negative about when it comes to inter-family communication. Certainly for the unhappy spouse, the abused child, or the isolated individual, online communication is more beneficial and positive than would be much of their offline communication.

The validity of the Carnegie Melon study, which suggests that by using the Internet people may reduce their social involvement and see declines in their psychological well-being, may be challenged on a number of important aspects. The findings of the study are unreliable, and cannot be generalised because they: 1) use a sample population consisting of people with little or no prior access to personal computers or the Internet and who have a much different motivation for going online than most users, 2) rely upon questionnaire data collected on only two occasions, 3) base their conclusions upon a insignificant (1%) increase in self reported levels of loneliness and isolation among respondents, and 4) value too heavily, and assume the positivity of, offline communication (especially within families) in comparison to online communication. Howard Rheingold concludes that, "Now that I have read the research report ... I can't say that HomeNet offers anything resembling significant evidence that cyberspace is undermining civil society." (Rheingold, 1998)

The findings of my own research of AOL users stands in sharp contrast to the above critiques, including the HomeNet study. In this research, I have spent over 200 hours over a 4 month period conducting participant observation on AOL. Data has also been collected through the use of email questionnaires and online interviews. Ideally, those conducting social scientific research in cyberspace would conduct follow-up face to face or telephone interviews to verify some of the self-reported data online (Hine, 1998). In my case, however, limited time and financial resources have made this impossible.

During this research, I have spoken with approximately 100 AOL users. All respondents have given their informed consent in exchange for guarantees that steps are taken to protect their anonymity. I've sought out members of AOL who are willing to disclose their feelings about community, online community, and friendship. I've done this by spending time hanging out in chat rooms, reading posts to public message boards, and seeking out people using the members directory and later contacting those people using private messages. These users were individually selected by the researcher to come from a wide range of demographic backgrounds and to reflect various levels of experience using computer networks. This sampling method, frequently employed by ethnographers such as myself, does introduce researcher bias to the study. The introduction of such bias is generally viewed as acceptable by social scientists conducting exploratory research where there is little known about the sample population as is the case here. It is usually unwise to generalise the findings of research conducted in such a way, but the findings do offer a valuable and much needed starting point for later studies.

In earlier sections of this article, the term and concept of "community" is discussed at length. Barry Wellman, a researcher at the University of Toronto, is one of the first to note the current shift away from communities based upon geographic areas towards private network communities. Wellman writes that today, "Instead of gathering in parks, cafes, streetcorners with easily-accessible neighbours, people now chat with friends by electronic mail and telephone or they get together at small gatherings in private homes." (Wellman, 1995, 1) Wellman calls this the domestication and privatisation of community. Fordism has divided labour and labourers, and zoning restrictions have created a situation where many must travel long distances to work. (Wellman, 1995, 1) When we travel to work, many of us must do so as individuals, and there becomes a decreased chance of unplanned meetings since we spend so little time in public areas. Unused public areas often become privatised, and as the public space in which community members may freely associate disappears, community life has also increasingly become privatised. Wellman writes that:

Additionally, Wellman notes that the home of the average person, especially within the suburbs, is larger today than ever before. This makes the private home an ideal place to host small, invitation only gatherings.

Wellman is not alone in noting the loss of public community spaces. Ray Oldenburg writes of the decline in publicly accessible space for informal association amongst community members to occur. He calls these places "the great good place" because, in his view, such places are "the core settings of informal public life". (Oldenburg, 1997, 16) The "great good place" exists as a third place in the lives of community members, following the first place (home) and the second place (work). According to Oldenburg, the existence of the great good place, or third place, is necessary for the good health and strength of communities. Without such third places, interaction between community members declines and the common ties shared by members of the community go largely unknown. (Oldenburg, 1997, xxiii & 72)

Oldenburg notes that the number of third places within cities throughout the Western world have been in decline for quite some time and that that this phenomenon is especially pronounced in the United States. In cities throughout the USA, the population of the suburbs is growing rapidly while the population of the inner city remains fairly constant or is even in decline. Suburb dwellers usually don't live within easy walking distance to places traditionally used as third places such as neighbourhood shops, a local pub, or coffee shop. Oldenburg writes that, "Houses alone do not a community make, and the typical subdivision proved hostile to the emergence of any structure or space utilisation beyond the uniform houses and streets that characterised it." (Oldenburg, 1997, 4) Or, as Richard Goodwin complains, in the suburbs "there is virtually no place where neighbors can anticipate unplanned meetings - no pub or corner store or park." (Goodwin, 1974, 38) It has even been demonstrated that the architecture of our cities increasingly discourages free association amongst members of the community (Davis, 1992).

Wellman also notes that fewer people are bothering to visit the few third places which are still available to them. He writes:

"annual movie attendance decreased in Canada from 18.2 times per year in 1952 to 3.2 times in 1987. Suburban shopping malls in no way resemble Atheninan agoras or pre-industrial public squares. They encourage fast-turnover consumption and discourage leisurely discussion. Unlike the public space of previous centuries, these privately-owned, profit-oriented spaces sell goods primarily for private, domestic use." (Wellman, 1995, 1)

For Oldenburg and Goodwin, as community spaces disappear, the communities which once utilised them also go into decline. What Oldenburg, Goodwin, and countless others are actually suggesting is that there is a noticeable loss of community feeling within neighbourhoods where the third place has become absent.

This does not, however, necessarily mean that community no longer exists. In his research of social networks, Wellman goes looking for the people who no longer inhabit third places, and finds that instead of disappearing completely, interaction amongst community members has shifted away from physical space into spaces created by new technologies. "Privatization means that people must actively contact community members to remain in touch rather than visiting public places and waiting for friends and acquaintances to pass by." In later paragraphs, Wellman explains the findings of his own research group:

Earlier in this article, we read several theoretical critiques about possible negative affects of computer mediated communication upon community. For example, Heim is quoted as saying, "Technology increasingly eliminates direct human interdependence. While our devices give us greater personal autonomy, at the same time they disrupt the familiar networks of direct association." (Heim, 1993, 100) We also read a similarly negative critique by Kroker and Weinstein who believe that computer network users are becoming "lonely telematic individuals huddled around terminal event-scenes (computer screens, TV sets, high-performance stereos) willing themselves to become members of a virtual community..." (Kroker and Weinstein, 1994, 39) The work of Wellman, on the other hand, stands juxtaposed to these critiques of computer mediated communication. For him communities based primarily upon third places and neighbourhoods are a thing of the past, if they ever existed at all. Communities continue to exist, but are supported through a number of technologies including the printed word, transportation, and new communications technologies. Computer mediated communication is just one of the many technologies used by people within existing communities to communicate, and thus to maintain those community ties over distance.

In this study, I use individual respondents as the starting point of network communities rather than looking at the network communities themselves. Because of this, when I talk of network communities, I am speaking of the private network community of an individual. This privatised version of the network community may or may not be shared, in it's entirety, with others. For example, my personal network community is made up of members of my own family, people I only know through email and other forms of online communication, people I've met on my travels, and people who I have come to know at various universities and conferences I've attended. We are a group of people and I am the common tie between us, although in many cases the interests of members of my personal network community do overlap. Some people within my network community interact with each other directly, while others do so exclusively through me. Our interaction takes place in a shared area, although this area tends to be electronically created. Communication between us takes place mostly through electronic communications technologies such as the telephone and the Internet although there are members of my network community with whom I regularly meet face to face. This is my own private network community and there isn't anyone anywhere in the World who shares this same exact community with me, although there are some who share large sections of it.

The concept of private network communities, based upon the individual, is not inconsistent with the definition of presented in earlier sections of this article. Network communities consist of a group of people. Members of the community share at least one common tie, even if it is only that they know an individual within the network community. They also interact with each other, although sometimes this is done indirectly through an intermediary. Within network communities, the area shared by community members may be offline and tangible, or it may be electronically created space such as those found between telephones or online. For many people who are part of network communities, face to face meetings are rare, so they spend time associating with each other and community building in electronically created spaces.

In the longer paper upon which this article is based, the ethnographic stories of a number of individuals are presented. Due to restrictions of the present format, however, it is only possible for me to comment more generally upon the findings of this research. Nearly all of the AOL users who have taken part in this study were initially motivated to obtain an AOL account by the need to conduct research online for academic or business purposes and to communicate with members of their pre-existing network communities. Not a single respondent, out of a sample consisting of just over 100 users, reports that they were motivated to join AOL with the intention of building new friendships there.

Not only were respondents motivated to join AOL by the need to conduct research online or to communicate with members of their pre-existing network communities, these tasks continue to be their primary activities even after several months, or even years, online. As continuing users, only 10% report that they ever seek to build new friendships online using their AOL account and this remains only a secondary motivation in their continued use of the service, after research and communication with members of their pre-existing social networks.

The majority, 75%, of respondents report that their private social networks are healthy, and feel positively about the number and strength of their interpersonal relationships. 19% of respondents agree with the statement that they "sometimes feel socially isolated", 5% report data in ways which make it impossible to determine whether they ever feel socially isolated or not, and 1% report that they "always feel socially isolated or alone". In reference to the 19% above, it is entirely natural for most people to feel socially isolated at one time or another in our lives, so this finding is not at all surprising or troubling. Every single respondent in this study feels that using AOL has helped them to maintain pre-existing offline relationships. In fact, I have found no evidence that using AOL harms offline social networks nor that it leads to social isolation for certain users. If anything, computer mediated communication through networks such as AOL or the Internet, appears to help people stay a part of a network community despite barriers of distance and time. However, when a person decides to spend time online, whether it is for work or leisure, it cuts into the amount of time which that person can use for other purposes. In the case of employees using the Internet, many businesses have recognised this and are limiting the amount of time employees can spend online. When a person spends their leisure time online, it similarly cuts into the time they have to do other leisure activities offline. For 97% of the respondents to this study, online time is spent at the expense of television viewing or reading. For them, each hour spent online during leisure time means one hour less spent watching television or reading. Much of this time spent online is used to send emails to members of user's pre-existing network communities, to participate in online chats with these people, or to conduct research.

AOL users who have taken part in this study are not socially isolated individuals desperately seeking contact as some would have us believe. They are not motivated by social isolation or loneliness, but by the ease with which they can access information online and keep in touch with their friends, colleagues and relatives from offline who increasingly have email accounts and web pages themselves. For some of these people, distance is an issue since they may find that members of their social circle have moved far away, but this needn't be the end of that social circle. Many who research cyberspace write about the ease with which barriers of distance are overcome by computer networks, and this is never more apparent than when members of a network community are able to communicate when they are located thousands of miles apart.

For the very few respondents who do report that they are socially isolated, most of whom report that it is only some of the time that they feel this way, being online gives them the hope of making new offline contacts and friendships which will continue into their offline lives. This is an entirely realistic hope: many members of online communities attend what are often called "flesh meets" where they get together with people they may or may not previously have known face to face. Some users even make bonds so strong with other users online that they go into business together or get married offline (Case, 1997, 2) In other words, even for the minority of respondents who do report that they sometimes feel socially isolated, using computer networks is a beneficial use of time.

For those who have access to computer networks such as AOL, CMC can be an efficient way to communicate with members of their pre-existing network communities. Although CMC can be used by those seeking new contacts and friendships online, it appears that such a use of computer networks is more rare than once thought.

Geertz, Clifford. (1973) The Interpretation of Cultures: Selected Essays. New York: Basic Books, Inc.