MAS963: Analysis of Interactions in a Newsgroup

soc.motss and me

soc.motss is a social Usenet newsgroup for the discussion of issues related to gays and lesbians. It is one of the oldest social newsgroups on the Usenet, with a continuous history of almost 15 years. In that period it has turned into a very clubby, very significant social place for its members. The soc.motss FAQ gives one introduction into some of the motss culture.

I have been a participant in motss for eight years. The newsgroup has been an integral part of my life, at times occupying two hours a day of reading, writing posts, and emailing other members. This interaction has spilled over quite a bit into "real life" - I have met roughly 200 people from the newsgroup, most of my friends outside of professional contacts are from motss, I even met my partner through a motss connection.

conversations

At first glance, a "conversation" in motss seems to be exactly a thread. Someone will post a message, people will reply, reply to the replies, etc, creating a thread. The thread structure is useful, but in practice it doesn't quite work. The conversation often changes topic in midthread. And one thread will often change into several different topics - flirtation between two people in one subthread, a serious discussion of political implications in another, and a series of cascading puns in a third.

Current threading newsreaders don't make separating subthreads very simple. The way I manage choosing what I read is to note the authors. Not only do certain authors have reputations for writing certain kinds of articles, but certain groups of authors are known to interact in certain ways. For example, if some particular pair of people are replying back and forth to each other rapidly, I can tell that it's yet another version of the same tiresome debate they have been having for a long time. Similarly, there are groups of four or five people who delight in making inane puns back and forth - I know to skip a long subthread with only their names on it.

At times there are conversations that transcend thread boundaries. Points made in one thread can pop up in another. This happens relatively seldom, though, and is often a bit confusing.

Threaded newsreaders became common in roughly 1991. An interesting comparison to follow up is to try to determine whether the introduction of this technology changed the feeling of motss much. I remember myself when I switched to a threaded newsreader. Before, I would read every single post to motss (roughly 100 a day at the time). This would mean that (because of time ordering) I would often read several unrelated posts by the same person in a row. With threading now I pick and choose and only read roughly 20% of the newsgroup (which averages 200 posts a day now), and my reading experience is shaped much more now by topic than author. I've had to redevelop the ability to construct a reputation for an author.

community

motss is a quite tight-knit community. There is a distinct feeling that we are all together, and even "bitter enemies" often feel a strong camaraderie. This comes partly from the subculture - gays and lesbians are a minority in society, there's a strong need to stick together. It also comes from the particular nature of motss itself.

motss discussion varies all over the map. The conversation is by no means limited to gay and lesbian issues - most every topic has been discussed, although there's certainly a somewhat queer spin on things. Much of the discussion is high levity and silliness - "cocktail party chatter" is the metaphor motssers themselves use.

But motss has always been a supportive place for people to talk about the difficult aspects of being a sexual outsider. There have been numerous cases of young people coming online saying "I think I'm abnormal, help!", older people discussing the problems of being single, mutual grieving and support about AIDS, etc. The belief that underlying all this silliness there are serious issues of support and subculture helps focus the community.

One particular feature of motss that makes the newsgroup more close is that we have yearly "motss.cons", weekend long parties where people travel to a host city and hang out. (I've been to two.) The attendance at a con is relatively small, maybe 70 people at most. The experience is always incredibly warm and social. Everyone already "knows" each other, so there is much mutual context, but at the same time it's incredibly exciting and new. The cons have helped cement friendships for much of the membership of the group. One common experience is for people who used to a argue about everything online to meet, get to know each other, and then understand each other's posts better now that they have more of a sense of voice.

cabals, "idjits", and taboos

As a participant of motss, I feel like there are roughly 500 people there. In reality there are many thousands of readers who lurk and never post. In 1993 when Usenet surveys were still being done, I believe the estimate was 80,000 soc.motss readers. (The number is almost definitely much larger now). These lurkers seem to have zero impact on the community. Some few communicate with public members via email, and a couple have shown up at the con, but in general they are zeros. Motss feels like a surprisingly small, clubby place for such a large readership.

One reason motss is so clubby is that the visible members (those who post) are a relatively small subset. Another is that motss members (in my opinion) deliberately enforce a community through private language, in-jokes, and a set of common standards.

The common standards enforcement is one of the most important, but also disturbing aspects of motss. There is a large unspoken consensus about particular political and social issue. For example, in American gay culture it's common to use feminine pronouns as a sort of gentle insult to a gay man, as in "what's her problem?". The motss cohort considers this sexist, and that view is enforced rather vigorously. Motss regulars know not to violate this particular taboo and or ones like it. When a newcomer comes in and unwittingly steps on one of these landmines, he or she is often taken to task about the remark. The way that the newcomer responds to this criticism is a test of their worthiness - a quick apology is always accepted, but people who argue back are often hounded until they either conform or leave. At times the word "idjit" has been a particular codeword for people who repeatedly violate the informal rules of appropriate behavior.

I believe that this set of behavior is a classic example of a community retaining its sense of identity in a larger community. The standards themselves are significant, but the enforcement of the standard is a way to keep the community together. motss has changed a lot since the Internet became a public phenomenon. To me, the group is less supportive, less emotional, less closely knit than it used to be. I now turn to private mailing lists for more meaningful interactions.

sketch

So this is a rather abstract sketch, not at all suitable for an interface. I was trying to capture two overlapping things about motss: the structure of the conversations (threads) and the structure of the community (people and their status).

The blocky stuff on the right is threads, rigid and boxy and nested hierarchically. The blobby stuff on the left is people. Posts are colour coded by person (not a very good visual device, I know). The shape of the blobbiness is intended to convey something about my notion of their personality. The larger the blob, the more influential or well known a person is. The density of the blobs reflects a person's interest - long spiky shreds would indicate someone who talks about scattered things, focussed soft round blobs are people who are more integrated, whole. Another interpretation might be that spiky blobs are mean people and round blobs are nice people.

This sort of divided picture is really quite unappealing to me. How can we integrate the rigid structure of a threaded conversation with the soft, subtle aspects of people?


This page is part of a group of pages for a class at the MIT Media Lab.
Nelson Minar Created: September 15, 1998
<nelson@media.mit.edu> Updated: December 15, 1998