For documents are much more than just a powerful means for structuring and navigating information space--important though that is. They are also a powerful resource for constructing and negotiating social space.
John Seely Brown and Paul Duguid [20]

Alt.good.morning (AGM) is a computer mediated group on Usenet; that is, the group uses computers and Usenet in creating and maintaining their systems of meaning. AGM is a newsgroup, a collection of messages and the group of people from around the world who wrote those messages. AGM is defined by the individuals who participate, the community they have formed, their purposes for participation and the computer systems on which they post. Without the computer, this group would not exist. In this chapter I describe the structures of Usenet that underlie the communication within a newsgroup. Additionally, I detail the structures of newsreaders by which posts are exchanged. Finally, I describe AGM, its participants and their participation practices.

Usenet is a global exchange of messages. It began as a discussion of Unix but quickly grew to thousands of discussions on thousands of topics. Although the early plan for Usenet estimated a maximum size of 100 sites with one or two messages per day [114], as of July, 1995, it is distributed to over 300,000 sites around the world with an average traffic load of over 127,000 messages per day [110] dispersed among more than 6000 newsgroups [7]. Although there is no way to determine how many people actually read Usenet, there are over 30,000,000 people with Usenet accounts [110]. There is no consensus on how big Usenet is, how many newsgroups there are or how many people read news, but everyone agrees it is big.

The most popular newsgroups are the social ones [12]. Network groups coalesce because of shared interests, not because of shared physical location. In his July, 1995, statistics on Usenet [110], Brian Reid reported that of the twenty groups with the highest number of potential readers, the discussion ranged from television shows to humor, sex and employment. Among the groups with the highest numbers of messages, over one third of the messages were about employment (50,078 messages in two newsgroups for 39% of the total traffic); another third, about politics (41,199 messages in three newsgroups for 32% of the total); and a fifth, in one newsgroup, rec.games.trading-cards (25,644 messages for 20% of the total). These particular topics contrast with those reported for March, 1993, in Baym [12] where social issues such as Indian culture, abortion, homosexuality and guns took up half the volume of messages and sex took up another fifth. Two years later, employment is an issue, the newsgroup rec.games.trading-cards is a dominant forum and newsgroups discussing sex account for only 15% of the traffic. Although the particular issues have changed, all in all, social issues still dominate the discussions.

A Usenet message is called an article or a post; sending an article to a newsgroup is called posting. Newsgroup participants post articles from their private computer accounts and they read posts using specialized software, either newsreaders or net browsers. Yet, most messages are not saved. This contributes to the perceived ephemerality of the communication. Posts are not spoken communication even though many threads (groups of messages) read like coffee klatsch gossip or a play, simulating conversation. Neither are many posts formal written documents. "In simultaneously bringing back lost arts of chatting and letter writing" [99], Usenet posts fuse oral and written communication [42, 99]. "Usenet groups are in many ways like oral cultures; without shared spaces or significant tangible objects, group tradition, coherence and meaning must be continually recreated through social interaction" [12].

Usenet is a mediating tool in all this message exchange. The spatial and temporal aspects of Usenet influence the identity, group coherence and group structure of a newsgroup [12]. Usenet is an asynchronous distributed system, that is, posters are geographically distributed and articles are received minutes, hours or even days after they are posted and have made their way among the sites around the world. Usenet is a public multicast where the conversation may be a monologue, a discussion or a multilogue. Posters do not know when their messages will be read or who will be reading their messages. Indeed, because most of the readers do not post, they are given the special label of lurkers. Participation in Usenet is therefore largely anonymous. Even without the guaranteed anonymity of lurking, a poster sits in front of the computer screen, alone. No matter how many people are connected to Usenet, "cyberspace is spacious. There is little sense of crowding" [99]. This combination of perceived invisibility and anonymity has been blamed for the lack of social norms and the increase of flaming [130]. In addition, messages do not arrive in topical order, a single thread may take days to complete and multiple threads are ongoing simultaneously. Reading messages is not the same experience for every person in the newsgroup. Because of the asynchronous nature and differences in delivery times of messages to some sites, messages and their followups are read in different sequences. These factors seem to doom a newsgroup to incoherent noise. Yet this isolation, disorganization and lack of consistency contrast with the connectedness felt when reading articles by people in your newsgroup.

Newsreaders, the software used to read and post articles to a newsgroup, are also mediating tools. Newsreaders format posts and allow posting similar to email tools. Thus, posts have headers, a message body with a quotation system and signature files. Additional organizational tools include kill files and threading facilities. Experienced participants use the organizational tools provided by newsreaders to structure their activities on Usenet.

The top lines of a post are called the header. Headers automatically accompany every post. Depending on the software used when the message is sent and what options are set for reading messages, a participant may see a rich or full header with detailed information about the post or some subset of those lines. A rich header includes the message number in this newsgroup at this site, the route through the sites, the sender of the message, the newsgroup(s) to which it has been sent, the subject, when the post was sent, the machine and organization of origin, the message length, its unique identification number, other posts referenced in the message, as well as other lines. The following is an example of a typical rich header:

Article: 41039 of alt.good.morning 
Path: news.tamu.edu!newshost.comco.com!news.texas.net!imci2!newsfeed.
internet.mci.com!news.sprintlink.net!gate.net!miafl2-42.gate.net!user
From: hp5997@bloom.tamu.edu (Holly Patterson) 
Newsgroups: alt.good.morning 
Subject: Re: It was a nice day 
Date: Sat, 30 Dec 1995 15:39:17 GMT 
Organization: AGM 
Lines: 17 
Distribution: world 
Message-ID: <4cl8ib$s0r@miafl2-42.gate.net> 
References: <649480980wnr@cream.of.wheat.edu>
Reply-To: hp5997@bloom.someplace.edu 
X-Newsreader: TIN [version 1.2 PL2] 
Content-Type: TEXT/PLAIN; charset=US-ASCII 

Headers are a rich source of contextualization. The article, from, subject, and organization lines are "most often used to establish participant structure and topical coherence" [12]. Each of these lines provides more information than what is displayed. Much can be inferred from these lines.

The article line indicates the sequence number of the post in the named newsgroup which have arrived at this site. In the example above, this article is number 41,039 of alt.good.morning to arrive at the Texas A\&M University site. This is a moderate number (only five digits instead of six), indicating neither a very prolific newsgroup nor one with low participation. Therefore, a newbie (novice) may consider the volume of this group not so high as to be unmanageable yet high enough to be interesting. There is only a moderate communicative history; most topics will not have been exhausted. At the same time, there may not be a well-established clique of participants; novices may be welcomed in such groups and their inadvertent blunders may be better tolerated. High volume groups may have more social norms and more established personalities [12} with the attendant expectation that newbies have read the FAQ and lurked long enough to know the rules.

The from line provides readers with the email address of the poster. It may also contain a name or pseudonym of the poster. When a name is included on the from line, either the email address or the name is set off in parentheses or a set of angle brackets (< >). Because these are unique identifiers, email addresses become one source of identity for newsgroup participants and help negate anonymity. Because country codes and other identifying information are coded into the email address, it often gives the location of the poster. In the example, Holly has a computer account at an educational facility in the United States. This indicates that she is either a student, faculty or staff member at the school whose initials are "tamu." Furthermore, since the from line identifies the author of a post, it can be used to filter messages from specific posters.

The subject line is used to indicate the topic of the post. It is provided by either the author of the article at the time it is posted or by the newsreader as in the above example. When a poster replies to an article, the newsreader software inserts a Re: in front of the previously supplied subject. This message is part of a thread called "It was a nice day." Although respondents can modify the subject which would begin a new thread, the subject line provides a mechanism for organizing posts in topical instead of chronological order which aids the coherence of communication. The subject line is also used to select or deselect which messages to read.

The organization line provides another identifier. Some systems automatically provide an organization name while others provide a way to customize this line. In the above example, because Holly has changed the default value in this line to read "AGM," she is affiliating herself closely with the newsgroup rather than the institution from which she posts.

The header provides social contextualization about authors and their messages which helps create and maintain internal organization. Another source of organization is the quotation system used by newsreaders. When you respond to a post, the newsreader includes the text of the original message offset with an angle bracket (or sometimes a colon) preceded with identification information. Some newsreaders include the date, time, subject of the quoted message and the email address of the author given in the original from line:

On Dec 30, 1995 15:39:17 in article <Re: It was a nice day>, 
'hp5997@bloom.tamu.edu (Holly Patterson)' wrote:

More often, the identification information contains the message id number and the email address of the author given in the original from line. You then insert your response at relevant points. Netiquette suggests including enough of the original message to be understood and that it is "extremely bad form" to include all the previous message [48]. Not withstanding the Netiquette norm, three AGMers make it a practice to include the entire previous post, usually after their few lines of response. Their practice is considered a public service by AGMers who routinely have trouble with inconsistent delivery of posts. Because the previous post is always included at the end of their followups, the included message is easily skipped if it has already been read. The following example illustrates the abbreviated message header and uses the quotation system:

In article <819988994.18094@bloom.tamu.edu>, 
'hp5997@bloom.tamu.edu (Holly Patterson)' wrote: 
> Besides, I'm a bigger addict to AGM than you are now! :)

*grins* 
I don't think there's any cure for you!  
Widge.....also addicted! :)

Embedding your remarks within a previous message is essential to maintaining conversational coherence. The original message provides contextualization by refreshing the memories of those who have previously read the post or by displaying the post for those who have not yet seen it. The quotation system provides the reader with a sense of turn-taking evident in spoken conversation. In addition, the angle bracket signals that the current poster is not accountable for the content of the quoted text but is accountable for the accuracy of the quotation [12].

Another common feature of a post is the signature (sig) file that is appended to the end of messages. Some systems automatically append the sig file to every message sent, thus providing a regular and consistent mechanism associating text with the author. Two hyphens often proceed the signature in these automatic systems, providing an additional visual cue to the reader. The sig file is highly individualized with ASCII art, a favorite quotation, address(es) and telephone number(s). Netiquette suggests that the sig file contain the sender's name and email address and be no more than four lines long [48]. Distinctive sig files carry more identifying potential than names and often provide the most salient cues in learning to discriminate between individuals [13].

Two functions of newsreaders provide readers with the ability to structure which posts they will read. Newsreaders provide a list of unread posts that are organized either by the time and date they were received or by the topic as indicated in the subject line. Messages that are arranged by topic are called threaded messages. A thread is a group of responses to an original message. The subject line usually has Re: and then the original subject; changes to that line cause subthreads to appear. As previously explained, the frequency of reading influences the efficacy of threading. The more often you read the newsgroup, the more sequential the experience even when posts are threaded. Continuity is aided by those newsreaders which renew the entire thread when new responses appear. Threading, therefore, provides a powerful method for constructing and maintaining conversational coherence as well as contextualizing the messages of a newsgroup.

The second function is the "kill file." You use a kill file to skip over messages [104]. Since you do not see them in the list of unread posts, the messages are effectively deleted (killed) from the list. You may kill posts that contain a specific subject line or come from a particular site, individual or newsgroup. You may kill entire threads, messages that are crossposted and messages containing a particular word or phrase. The kill file provides an efficient filtering device by which readers can structure their participation in a newsgroup.

Usenet provides an environment for discussion on a multitude of topics that captivates millions of people worldwide. At the same time, Usenet presents obstacles to and disorganizing influences on newsgroup communication. The spatio-temporal separation of participants hinders group cohesion and obscures individual identity. The massive quantities of messages interfere with group consistency and coherence. On the other hand, newsreader features provide structural support and organization which lessen the "chaotic influences of Usenet" [12]. AGM uses these features in creating its community.

Since AGM is a Usenet newsgroup, what holds for Usenet and newsreaders holds for AGM. The features of Usenet and newsreaders affect the participation on AGM, yet the group has found ways to use those features to build community. The following description of AGM is from three months of posts to the newsgroup, a detailed analysis of six weeks of those posts to AGM, data from an electronically distributed questionnaire, the file of frequently asked questions (FAQ) and interview data. In the excerpts from posts and emails, I have not corrected the grammar or spelling. This is the way they were sent; some mistakes are typos and some are mistakes made by people who do not normally speak English. Either way, these are the words as AGMers expressed them.

Alt.good.morning as the newsgroup name gives little indication of the purpose of the group or general topic of discussion. The description in the Newsgroup Info Center (NIC) is "Would you like coffee with that?" [7] and that provides even less information. More appropriate is the purpose from the perspective of the participants. As one participant says, "AGM is a forum for people of the world of a like nature to meet and become friends and support the ideals of caring and sharing, support and friendship." Other features of AGM include "friendliness, silliness, honesty, understanding & bounces!" AGM provides a means for creating friendships in a nonthreatening environment composed of acceptance, humor, support and trust. The following exchange between a lurker and an AGMer illustrates the purpose of AGM:

> as i reset my news reader, i must ask , what is the point of the group?

The point????  Caring, supporting, connecting, reaching out, silliness,
recipes and advise, a bright spot in the day, a port in the storm, a 
helping hand, a lifeline, questions,  answers, and the weather report! 
In a word, sharing.  Stick around, read and listen, jump in and be one 
of us...have a cup of warmth, a slice of life and a chair of bowlies.  
We are here and we extend a heart to you. 

AGM is newsgroup in the alt hierarchy; that is, it was created without having to go through any discussion or vote as do newsgroups in the major hierarchies [9]. AGM began in the winter of 1992-1993 as a small group of posters in Europe wishing each other "good morning" and has grown to over 80 messages a day from people all over the world. The conversation of AGM is routine and mundane with no special topic or purpose to bind its participants. Just as there is no agreement on the number of newsgroups, there is no agreement on the size of a newsgroup. The estimated number of AGM readers ranges from 6,100 [7] to 15,700 [110] ranking it 2,260th among Usenet groups in readership size [110]. Brian Reid further estimates that 43% of the Usenet sites receive this group.

A complete description of AGM cannot be constructed from one data source. It requires layers of descriptions from multiple sources and multiple perspectives. The first layer of description is based on a detailed analysis of posting patterns over six weeks. The next layer of description is based on responses to the questionnaire I submitted to the group and provides demographic information. The remaining chapters of this dissertation provide additional layers of description.

Over the three-month period from October 21, 1995 through January 31, 1996, I received 9,839 posts to AGM. It was not possible to automate the counting of the number of different posters during this time. Several families use a family computer account, that is, one account for the entire family, and reading the messages was the only way to determine who actually sent the post. Many AGMers use multiple accounts for posting and there were groups of people who would post from an individual's account during special meetings. In addition, one poster was able to modify the from line changing the email address that would appear on his posts. Therefore, I decided to hand-tally this information over a six-week period from October 21 through November 30, 1995. The following description is based on the analysis from that time.

During this time, AGM was included in a crossposting campaign by a Canadian poster, Ray Moore (pseudonym), who was concerned that Quebec would vote to secede from the Dominion. Ray also included AGM in his religious spam. (Crossposting is sending the same article to multiple newsgroups; spamming is a virulent term for crossposting.) AGM normally receives approximately 1% crossposting [110]. Yet crossposts reached a high of 18% of the total posts during the week before the Quebec vote and the week after the vote. On 30 October, crossposts reached an all time high of 36% (the effect on AGM of this volume of "outsider" posts is examined in the chapter on community). Although other people included AGM in their crossposting, Ray's campaigns resulted in the most traffic of this type on AGM.

There were 4,047 total posts to AGM from the end of October through November (see Table 7). The numbers in this table come from the posts that I received and do not reflect the total posts submitted to AGM. The AGM-type posts includes those articles by people who regularly post to AGM as well as messages from people who have not continued to post to AGM but whose messages were in the AGM spirit, i.e., friendly. The crossposts include the massively crossposted articles by Ray, the followups to his posts and any crossposts from AGM regulars even when the articles were AGM-like. Because several of the AGM participants also post in other newsgroups, there were a few AGM-type messages crossposted to one other group and are therefore included in both categories. These posts did not burgeon into offensive spamming but expired as normal threads. Commercial messages were not crossposted and are usually of the "hit-and-run" variety. The posts that were not classified include the flames that were not crossposted even if they originated from crossposts and the informational messages not generally considered commercials.

Table 7. Total Posts to AGM

     Type of Post   Number  (Percent) 
     --------------------------------
     AGM-type       3,537       (87) 
     Crossposted      497       (12)
     Commercial        13      (0.3)
     Not classified    34      (0.8) 

In determining the average length of a post to AGM, I again had to evaluate the type of post and ignore the length of the header before the value acquired meaning. Crossposts seemed to be longer than normal AGM posts. Because of textual embedding, followups seemed longer than original posts. Then there were some special AGM posts that were significantly longer than the usual posts to AGM. One is the FAQ, which is routinely posted to the newsgroup in two files; one, 264 lines long and the other, 220 lines long. The other group of special posts was a thread entitled Apologies because of its increasing length. Apologies was a word-wrap game played by the group which grew to over a thousand lines of post and which was finally split by the forwarding software into two files of over 600 lines each. Including even these longer files, all the crossposts and the followups, the average length of posts to AGM was 34 lines. To obtain a clearer picture of AGM posts, I computed several different averages. The average length of a crosspost was 38 lines. Including the FAQ and Apologies, the average length of an AGM post was 33 lines; excluding them, it was 30 lines. Not only did crossposts seem to be longer, they were longer. The average length of an original AGM post (i.e., not a followup and not the FAQ) was 24 lines. An original AGM post is approximately one screen long. In addition, 75% of all AGM posts are followups to previous posts and 61% of all AGM posts include a sig file. The average length of a sig on an AGM post is 6 lines; on a crosspost, 5 lines.

Although crossposts accounted for only twelve percent of the total posts during this time, they affected the conversation on AGM. One of the problems with crossposting is the amount of traffic generated. Only 8% of the crossposts were original articles, that is, unique messages with a new subject line; yet even that number is too high because Ray would post the same article over and over, each time with a minor modification and a new subject line. Besides increased traffic, another problem with crossposting is the number of flames generated. AGM has an explicit injunction against flaming. As seen in Table 8 (Note: line graph in the official version of the dissertation, charted here for the online text version), as the number of crossposts increased, so did the number of flames. The effects of crossposting on newsgroups is little understood and special attention needs to be given to this phenomenon.

Table 8. Crossposts vs. Flames

           Crossposts  Flames                 Crossposts  Flames 
           ------------------                 ------------------
October 21      6        0     November 11         23       5
        22      1        1              12          5       1
        23     17        3              13         16       9
        24      6        1              14          9       2
        25     24        6              15         12       4
        26     20        5              16         11       2
        27     41       10              17          1       0
        28     26        8              18          5       1
        29     13        6              19          1       0
        30     37       11              20          8       1
        31     21        6              21          8       2
November 1     14        3              22         19       4
         2     14        4              23          7       3
         3      5        0              24          3       2
         4      9        1              25          4       3
         5     14        1              26          5       1
         6      9        3              27          5       0
         7     11        1              28          3       1
         8     16        9              29          2       1
         9     13        4              30          1       0
        10     29       10

The above description has been about posts; the following posting patterns are about people. Just as the messages were divided among AGM-type posts and non-AGM posts, the people must be classified. If the person posted a message in the AGM spirit, then that person is considered an AGMer. Others (non-AGMers) are crossposters and spammers who posted or replied to the massively crossposted threads and never participated in AGM threads. There were posts from 427 different people, 203 AGMers and 224 Others. Thirty-four people used multiple accounts and twenty-six additional people had variations in the format of their account information on the from line.

The information in Table 9 shows the posting patterns for the crossposters. Because these numbers include some AGMers who crossposted, the total does not equal the number of Others given above. The obvious pattern of these posters is that most post only one message and are never heard from again. A few posted over several days. Only one, Ray, continued to crosspost; over thirty-nine days he posted 106 articles. One of the crossposters finally did join in the AGM thread, has followed the norms of the group and continues to post periodically.

Table 9. Posting Patterns of Crossposters

     Frequency        Number 
     -----------------------
     1 day, 1 message   163
     1 day, >1 message   22 
     2 days              29 
     3 days               9
     4 - 5 days           6
     6 - 9 days           4
     10 - 19 days         0
     20 - 29 days         0
     30 - 39 days         1

The information on posting patterns of AGMers is more difficult to classify (see Table 10). Because people come and go in a newsgroup, sometimes posting, sometimes lurking and sometimes not participating at all, posting pattern data is incomplete. For such cases, these people are not included in Table 10. In contrast to the crossposters, less than a quarter of the AGMers posted only one message. The high number of AGMers who post once a day is related to an explicit norm that states that you should post once a day. What is evident is that AGM has an active group of posters.

Table 10. Posting Patterns of AGMers
     Frequency        Number
     -----------------------
     1 day, 1 message     47 
     average 1 / week     10
     average 2 / week     14
     average 3 / week     10 
     average 4 / week      3
     average 5 / week      3
     average 6 / week      1
     average 1 / day      21
     average 2 / day       6
     average 3 / day       2
     average 4 / day       1
     average 7 / day       2

There were twenty-eight people who contributed forty or more messages during this period, fifteen women and thirteen men. Six of these contributed more than one hundred messages, three women and three men, including Ray. Two of these six contributed over two hundred messages. The twenty-two heavy posters contributed over 56% of the AGM-type messages and the five very heavy posters (excluding Ray) contributed over 28% of the AGM-type messages. These six heavy posters greatly influence the discourse on AGM.

But who are these people? To get a better idea of who AGMers are, I posted a questionnaire to the group after the three-month data collection period had ended. By that time, the massive crossposts had ended and AGM was back to normal. I received 112 responses. All five of the very heavy posters (Ray had disappeared by this time) responded as did one lurker (who now posts to the group). Not everyone answered every question and some provided multiple answers. Over half (54%) of the responses were from the United States and the other half fairly equally divided among the United Kingdom (15%), Canada (13%) and twelve other countries (16%).

AGMers range in age from under twelve to over seventy-one. The greatest number of posters are in their twenties (42%), university aged, which is not surprising considering ease of computer access at most universities. Another 21% are in their thirties and (coincidentally) another 21% are in their forties. The under twenty and over 50 groups are fairly evenly divided, 7% and 9% respectively. Of the 112 respondents, 57% are women and 43% are men. These numbers do not match the reported 70% men and 30% women on the Internet [106]. The group is well educated; 37% have an undergraduate university degree, an additional 17% have a master's degree, two have doctoral degrees and two have professional degrees. As expected considering the ages of the respondents, almost 30% said they were students. The other major categories of occupations include education (20%), government (10%) and computers (13%). Yet, other categories are represented, including retired, service, sales, arts and volunteer work.

AGM is a computer savvy group. Although 12% said they do not own a computer, 10% said they own three or more computers. Most access AGM from home (66%) and many also access from school (22%) or work (24%). The number of computer accounts also gives an indication of high computer usage among the group--half have one account, almost a quarter have two accounts and the other quarter have three or more accounts; this coincides with the number of people who used multiple accounts for posting during the six-weeks analysis period. Most of the people had been on the Internet from one to three years (58%) and most used the Internet for both communication and entertainment. Not unexpected, then, is that email is the most-used Internet application (60%) and Usenet (39%) is the second most-used application.

The next set of questions were about their perceptions of their access and affiliation to AGM. When asked how often they read AGM, 60% said daily, 13% said weekly and 16% said occasionally. When asked to classify their posting habits, 25% said they were regular posters and 62% said they were occasional posters; none considered themselves heavy posters, not even those who posted seven articles a day. The people who responded to the questionnaire considered themselves members of a group. When asked to classify their affiliation to AGM, several gave multiple answers; 51% consider themselves AGMers, 29% oldbies (a term for someone who has been on AGM for some time), 14% lurkers and 12% newbies. In addition, AGMers are serious about communication. Over 90% of the respondents communicate with other AGMers by posting on AGM (93%) and via email (94%); over half communicate via talker (60%) (a program similar to IRC), telephone (55%), FtF (54%) and snail mail (53%) (a term for the regular postal service).

AGM is a part of their lives. They talk about AGM with other AGMers (79%) and with people outside of AGM (74%). Newsgroups are still new to the general population and this is echoed in some of the remarks about talking with other people about AGM. Bette, a heavy poster who uses an abbreviated writing style, remarked that she talks to others "A bit but not much, it can be rather awkward and esp irl ppl often don't understand." (Her last phrase can be translated as "especially, IRL people don't understand," where IRL is a common acronym for "in real life.") Others mentioned that when they talk to their other friends about AGM "they think I'm very strange ;-)" and "they think it's a strange cult sometimes!!!!" Most feel that AGM has positively affected their lives. "It has given me a great community of friends." For Rita, a light poster, AGM is a family:

I feel like part of a large family, that means all the good and the bad. The sorrows, anger, joy are part of life and that people can express them without fear of rejection is so important. I spend more time now on my computer than before - is that bad, I don't think so. I'm learning more about people by being part of this newsgroup and hopefully will learn more about how to communicate, only with written words, which is new to me.

Family and friends are frequently mentioned in the context of AGM. For Charlie, a very light poster, AGM has had a more personal impact:

It has completely changed my life!!! Before I was lonely and lacked confidence and that has improved a lot. I had more confidence talking to people on a computer, and that has also taught me how to be confident irl (rather than making me a recluse at the computer all day!!). Many of my computer friends are very dear to me, and I finally fell in love when I was least expecting it, so yeah pretty cool :) :)

Those who talk with others about AGM and who feel that AGM has directly affected their lives are more likely to consider themselves addicted to AGM. Again, the words "family" and "friends" are used whenever AGMers talk about AGM. "I have made too many good friends to give them all up, now. These people and their lives are important to me . . . I like to feel that they feel the same. Giving up AGM would be losing a whole family of friends." For others, the word "addiction" is too strong--"I see addictions as something harmful, and I find AGM, very positive." Emily, a light poster, gives some of the reasons for the engagement in this newsgroup:

[F]or the first few months I definitely was [addicted]; the group was there to make me laugh, make me think and support me through some pretty tough times, whether they knew it or not. I logged on multiple times during the day and read everything. Since the irl meeting in AZ, my "addiction" has tapered off somewhat for a couple of reason; one, some of the "mystery" is gone and two, general life time constraints and the desire to focus on my life in my "real" community (the one where I live physically).

The group is diverse enough to be interesting to many different people from around the world and homogeneous enough to engender feelings of deep friendship. Mumford explained that people "tend to be more socialized at a distance, than they are in their immediate, limited, and local selves: their intercourse sometimes proceeds best, like barter among savage peoples, when neither group is visible to the other" [96]. Emily seems to lament the loss of that distance, what she calls "mystery," after her IRL meeting with some of the group, but she has only "tapered off" and not relinquished her computer life for her physical one.

Except for acknowledging the global nature of the group, AGMers seem not to be affected by the spatio-temporal separation of Usenet participation. It is there, but it is not a major factor in their communication. "Geography on the internet is simply a matter of server time - you can be with us at the click of a mouse" (Widge, very heavy poster). Perhaps this is also because they use a spectrum of communication media. Talkers, synchronous distributed media, are used by a large percentage of AGMers. One talker, Castle of Lost Dreams (COLD) is frequented by so many AGMers that it has been adopted as "the" AGM talker. In an email interview, Widge had this to say about COLD:

[T]he involvement of so many AGMers there makes for a very friendly environment for new people who come there. Many AGMers have come to the newsgroup via COLD because they first stumbled across us there and found it a sufficiently warming experience to be interested in the group that brought us all together in the first place. Sometimes, people switch to COLD almost exclusively because its a quick and easy 'fix' of AGMers! But they still think of themselves as AGMers even if they don't post or read the group anymore. I've had some discussions with Bette and others about this in AGM and about how much people who use COLD regularly, and effectively drop out of AGM, seem to confuse the two groups and assume that particpation in one place is the same as 'membership' of both.

AGM is real to AGMers, as real as a telephone call. These are real friends and real family. Two couples have met on AGM and married; two more AGM couples are now engaged. There is a variety of posting styles on AGM including the lurkers who post only occasionally, those who send original posts about their daily lives, others who use AGM for support during times of crisis and those who post often but seldom disclose much about themselves. AGM is "a place for the 3 F's: Fun, Feelings, and Friends" (Jan, very light poster).

AGM uses the affordances of Usenet provided specifically by newsreader software to create a group of friends. They are people of all ages and all walks of life. They access AGM from school, work and home. They communicate with each other often and with a variety of media. They strongly identify with the group and express feelings of deep involvement with and even addiction to the newsgroup. AGM is a part of their lives. The next chapter describes the community that AGM has built.