I'm going for a first-time event here: a triple cross-post at Easily Distracted, Cliopatria and Terra Nova. I'm at a meeting on law and virtual worlds at the New York Law School, and there's a really interesting panel discussion of methodologies in virtual worlds. Douglas Thomas just pointed out that when we talk about qualitative methods in virtual world research, we always tend to define that as ethnography, when there are other kinds of qualitative methods that are potentially important, including history.
I think that's right, and it struck me how odd it is that I, as a historian, generally talk about virtual worlds methodology in terms of my habitual dissatisfaction with the tendency of anthropology to visit its own ethical obsessions on all discussions of ethnography as a method. I don't talk about historical narratives or events in virtual worlds, even though what I think is most interesting about virtual worlds is that they are historical, processual, dynamic, iterative.
So what are the methodological challenges of tracing events and processes over time in virtual worlds? Well, part of the problem is that some of the richest quantitative or empirical kinds of data imaginable have been until very recently held privately by developers. Star Wars: Galaxies extensively tracked the history of the construction of housing within their gameworld, but without access to the developer's data, the only systematic conclusions you can make about that process depend upon personal (or pooled) observations and reports. A historian of virtual worlds to date would need to have been there to say much of anything about many of the structural histories of importance.
But this is even true for a narrative of events within games. If a 25-year old graduate student came to me and said, "I want to write about the history of events within Meridian 59 and Ultima Online, about the narrative evolution of the games, about key episodic things that happened", I'd pretty much say that this graduate student is in a worse situation than a historian of modern Africa. The textual sources are going to be extremely difficult to recover in a thorough way because there are both too many and too few; a lot of the rest will only be knowable through oral historical work, or through questioning people through email. I know about Dread Lord Days, and about the way Arwic was in Asheron's Call before secure trading was introduced, and about about Sunny in LambdaMOO and so on. I know about the fetuspult in Dawn. I was there for it all. So I could write that history as an eyewitness, from the perspective of experience. Of course, I could add some sustained archival research, because I know to use keywords like "fetuspult" and "Arwic".
I had a discussion earlier today about a parallel problem in simulations of emergent phenomena, which also seem deeply historical and processual by their nature. It seems to me that knowledge production around such simulations often requires experience, you have to watch a simulation again and again to begin to understand the range of variability in its evolution over time.
So I'm powerfully convinced that the history of any given virtual world, and the history of all virtual worlds, is a crucial part of knowing them in qualitative sense. But I'm also struck that this is actually harder to know than the already-difficult methodological challenges of my major field of specialization, African history, and for various reasons is also harder to know than the general history of online and new media, which are archived in ways that experience and events in virtual worlds is not.