First posted to the Sarai Reader List: http://mail.sarai.net/pipermail/reader-list/2005-January/004829.html

Hello, this is my first post here.

I’m studying how user interface affects discussion and the resulting community in online spaces. In keeping with the spirit of the investigation, I’ve posted the full text of the proposal to an online discussion space, my journal: http://jace.livejournal.com/351443.html

Ignore the comments, please. They make it sound like I won a medal. ;-) My original idea for this investigation came from LiveJournal itself, where I’ve been a regular user for several years and noticed usage patterns related to the user interface. An example:

LiveJournal supports threaded discussions—that is, a discussion where you can reply to anybody, the person who made the post or someone who made a comment on the post—a feature unremarkable in mailing lists because it’s taken for granted, but unusual for a blog, where the norm is that readers can comment only on the original post, not on another comment. LiveJournal goes one step further: when you reply to someone, they get an email notification with a copy of your comment, and links leading back to the comment online or to the entire post.

Combine threaded discussions with email notifications, and you have the ability to let your readers talk to each other without involving you. Here is a recent example from my journal: http://jace.livejournal.com/360412.html

Notice that my post of a picture of a clothesline has spawned an unrelated discussion on henna. From personal knowledge, two of the three people involved were not previously familiar with each other. Thanks to LiveJournal’s user interface, my post provided a space for these two to get familiar with each other, from where they will go on to engage in each other’s journals, essentially building a community where most people know each other well enough to be comfortable in their company.

Some people will argue that threaded discussions are bad precisely because they allow people to fork off into parallel discussions, taking attention away from the original post. While this may be a problem on Usenet or mailing lists, it’s not on LiveJournal because posts have a limited lifetime. This used to be a few days earlier, but it’s barely 24 hours these days (a sign of LJ’s UI design approaching overload). I will attempt to explain this in my next post to this list.

In around August/September last, when I first presented the idea for this investigation to the good folks at Sarai, but didn’t know exactly what I wanted to do, Joel Spolsky of JoelOnSoftware.com wrote an article on social interfaces in software: http://www.joelonsoftware.com/articles/NotJustUsability.html

His article helped considerably with clarifying the purpose of my investigation, and explains it better than I have here.

I’ve had a busy January honouring earlier commitments (and hence the delay of this post). During February, I will document UI related observations from various communities online. Since these observations have no backing other than my own claims, between March and July I will attempt to experimentally verify them with a willing community.

I’m looking for such a community. It will be a difficult search since they must be capable of modifying their software—and I don’t have the resources to build an experimental community just for this investigation—but I hope I find one. Please let me know if you are interested.

Have a good week, everyone.