Milwaukee.
The next speaker at this AoIR 2009 blog research session is incoming AoIR Veep Alex Halavais, presenting a paper co-authored with Helen Martin. He begins, though, by referencing previous work of a researcher recording graffiti in New York City: this was done as a way of tracing how people make use of emerging workarounds. In the 1970s, bathroom wall graffiti was the equivalent to what is now blogging, Alex says, so us bloggers today are essentially writing on bathroom walls. Tracking this provides a trace of how people work around their lack of access to participation and voice in the mainstream media, of how they manage to make their ideas heard regardless.
Today, there is a growing understanding that there are a variety of different blogospheres in existence, and a wide range of blogging practices; these must be studied as distinct types with their own rules and principles. Alex's study examines especially the top-level, high-traffic blogs, and here is interested especially in the practice of how on such blogs (with their extensive discussions in comments) the practice of addressing other commenters using the '@' sign has emerged ('@so-and-so, I disagree...'). Commenters themselves seem to think that such practices predate blogs, but there is little evidence for this - so we should be able to trace back where this started.
The study backtracked through a number of major blogs over the years to see where the first instances of the practice can be found, and saw it first on the Engadget blog; it was picked up fastest by technology blogs, and saw its heaviest uses on gaming blogs. The study was also able to identify a number of users operating across a number of sites who spread the practice, but it is still looking for its 'patient zero' who started the practice (and should at least in theory be able to be found).
Today, of course, this practice has been adopted (and formalised) in a big way in Twitter, and from there has re-invigorated use in blogs again, too, and it will be interesting to track it further. (In comments, Lois Scheidt suggests that the practice may have started on an IRC client, by the way...)