blogging and communities

I recently posted a comment on Feld Thoughts. (The blog's moderated. I don't know if he'll approve the comment, so you might not see it there if you follow the link.) The comment contained a few thoughts about blogs and online communities. I wanted to expand those thoughts a bit, and what better place to think than out loud, on the fly, and in front of the entire Internet? :-)

The gist of the comment was this: blogs have some advantages for forming communities, but they're not ideal. On the plus side, they provide content that acts as fodder for conversations. As word about the blog's theme spreads, people from around the Internet may gather there, so you have a natural nucleus for a community of people with a common set of interests.

They also have some drawbacks. Communities form despite the drawbacks, but if your goal is to build online communities, some design changes may be in order.

1. Blogs create conversations with unequal power. The blogger, the person who owns the blog, has the blog's full typographical facilities available, the ability to post pictures, and, most importantly, the ability to moderate comments. People writing in the comment space have a much more limited set of abilities. Now, sometimes that's fine. In fact, sometimes having a good moderator can improve the conversation. But more often it seems to impede the conversation and flow of ideas. Of course, if the comments section supports hyperlinks, the comment writer can always link off-site, but that leads to problem #2.
2. Conversations become fragmented over space. Suppose you have ten friends. Each friend has his or her own blog. Now you have the potential for conversations to sprawl through the comments sections of ten different blogs (times the number of active posts--see #3.) If each blog supports an RSS feed for the comments, you will at least have a notification that something new is there, but it still takes effort to piece it all together. Also, if a comment writer links off-site, people may follow that link, scattering them further.
   
3. Conversations become fragmented over time. Conversation threads on an e-mail list tend to rise and fall depending on peoples' interest in following them. With a blog, though, conversations tend to shift with each new blog post. So when the blogger puts up a new post, that's where most of the energy goes, and people who want to continue a previous conversation are likely to be chastised for posting off topic. It's true that there's really nothing to stop you from continuing to comment on an old post, but it'll get very little attention, especially as the post scrolls off the RSS feed.

So, how do you deal with these problems? #3 seems to be the easiest to tackle, at least in isolation. One way might be to have the comments section be a sort of continuous, threaded forum that wheels along on its own rather than seeing a sharp cut-off every time there's a new post. That's as opposed to today's design where each new post essentially creates its own new forum space. If you wanted to keep a connection between the post and the forum, it might work to have each new post automatically open a new thread in the forum.

Problems #1 and #2 seem interrelated. If you could solve the problem of keeping a coherent conversation across web pages, then you could give each participant his or her own web page, and the distinction between comment writer and blog author might disappear. It would require a few things.

1. Instead of reading blogs, you'd read authors: you would add authors to your client software instead of adding web sites to it. A community might consist of a centrally-maintained list of authors, much as we have central e-mail lists today, that you could add or drop as you choose.
2. Your reading software would have the job of locating your authors' comments and posts from around the Internet and collating them into some sort of meaningful conversation structure.

So have I just re-invented the publicly-archived e-mail discussion list with a killfile? Sort of. One difference is that it's a pull design rather than a push design--instead of a server sending e-mail to me, my software goes and finds the entries I want. That means I have complete control over who I do and don't read, decreasing the chance of my e-mail address drifting into a bulk spam list. Another difference is that a properly designed system would let people comment on web sites that don't have their own comment feature.

There's another problem I haven't addressed, yet: permanence. Stuff on the Internet tends to be indexed, archived, and searchable for a very long time after it's written. Sometimes that's great. Other times, maybe I want to kick ideas around without their being tied back to me because I'm just exploring and haven't really formulated my thoughts. For instance, consider someone just trying to understand what communism was all about in 1933, discussing its pros and cons. Those words could easily come back to haunt that person years later before the House Un-American Activites Committee. One way to deal with it would be to use pseudonyms, so an author could abandon a pseudonym every so often. There might be a better option, but I'm not yet sure what it'd be, since once you turn something into bits it's very hard to make sure those bits later vanish.