Mathemagenic

Together with Aldo de Moor we are working on analysis of weblog conversations. These are my "thoughts in progress".

In my paper on weblogs and knowledge work I distinguish between two types of weblog conversations: those in comments to a post and the ones "distributed" between weblogs. I guess in most of the cases both types are integrated into a whole, but this distinction is useful to look at mechanics of weblog conversations.

Starting point: technology. At a lowest level there are two ways to react to an earlier argument: using post comments (if initial argument is a weblog post) or commenting in another weblog.

1. Argument (weblog post) -> response using comments

• connection is symmetrical in most of the cases: it's relatively easy to find responses to an argument and vice versa
• visibility
  • visible while reading in browser
  • not necessary visible while reading using RSS reader
• this part of weblog conversations is not much different from a forum discussion or comments that follow an article

2. Argument -> response in a weblog (response post)

• connection is asymmetrical
  • backwards connection usually exists - response post links to initial argument in most cases (but this is up to the good will and permalink location skills of a response author)
  • direct connection exists sometimes: argument is linked to reaction post via trackbacks
    • this works only if argument is a weblog post and both weblogs are trackback enabled
    • doesn't work is argument is a general web-page or comment in another weblog
• visibility
  • backwards connection is visible
  • if exists direct connection
    • visible while reading in browser (but reader has to know meaning of a trackback)
    • not necessary visible while reading using RSS reader

To simplify things talking about visibility I refer only to things that you can see in a weblog itself and not using other tools (e.g. Technorati, Blogdex, referrer logs, search on a post title/keywords).

Characteristics of weblog conversations. As a result of technology used for weblog conversations, they are likely to be:

• fragmented and distributed between several weblogs
• difficult to trace 
  • need to use multiple tools to do so
  • never sure that everything was found
  • easier to trace backwards (starting from the last post), but then there is a risk of missing branches

Effects

Participants of a conversation are likely to have very different views on it (as it's not likely that everyone will trace all previous contributions).

As weblog conversations are not "locked" in one space they are likely to include contributions from different groups of people (as the opposite from a forum conversation, which is usually "within a group").

Because weblog conversations are difficult to trace, the participants tend to write self-standing pieces and often try to summarise (or at least link) previous arguments making it easier for a reader.

Weblog conversations are often compared with academic discussions in papers (of course, they are slower and tend to stay within borders of a discipline). I'm thinking of a better metaphor...

I think of water flows in a river delta: different streams join and separate again, mix, go faster or slower... Some of them are mainstream, large enough to follow by boat, while others are tiny and hidden, so only natives can find and follow them.

Weblog conversations are like that: a complex system of streams that feed the sea of ideas.

This post also appears on channel weblog research

Later: the paper I was writing about - An argumentation analysis of weblog conversations

It feels good then plans are turning back into reality: BlogWalk is getting more real. 

BlogWalk is a series of face-to-face meetings aimed to bring together weblog researchers and practitioners for in-depth conversations about their work.

The intention is to complement BlogTalk with smaller scale opportunities to meet and to talk. The idea was in the air for a few months already, but we couldn't start working on it. Then one late evening Sebastian Fiedler came up with this great name and it went rolling. Walk means that we will be "walking" around places and have fun there. It also means that meetings will include walking in and out of pubs and nice places to eat because everyone knows that alcohol and good food fuel conversations. Note: this fun side is about complementing hard reflective work together and not replacing it :)

BlogWalk 1.0 will be focused on weblogs in KM context and will be held on 19 March 2003 in Telematica Instituut, Enschede, Netherlands (this is where I work :). I'm looking forward for this opportunity to meet KM bloggers in a real life next to our virtual conversations.

And, a couple of days before Sebastian Fiedler and me will present on weblogs and learning in Utrecht University. Hopefully, we will also find time to work on a paper on weblog apprenticeship :)