I’ve been following the conversation between Elijah and Liz with some interest, but with little comment. Since my name is invoked in both laudatory and mildly disparagingly ways, I suppose I ought to respond. Despite disagreeing with much of what both of them have to say, I think the conversation highlights some critical issues.
*What’s a weblog? Who cares.*
I didn’t participate when Liz made an effort to start a conversation about “what a blog is,” because I didn’t find creating such a definition of much use. Liz criticizes Elijah’s (et al) panel because of his generalization of the _blogger_ which she likens to an attempt to discover the average _author_ of a book.
I was also in attendance at that presentation, and I shared some of that dismay. Chief among these was that the presenters were setting up a straw man in terms of the way most people thought about blogs. They led off with a description of blogs that did not at all fit either what I sensed from mass media coverage, nor what many people in the blogging community thought bloggers were. That is, the _problematique_ engaged by their work is that popular perceptions of blogs differ from empirical observations. I think that such a difference was overplayed, and the rhetorical presentation of the data in this way hurt rather than helped frame their descriptive work.
Especially problematic was when members of the panel made off-handed comments about “the average blogger.” Certainly, we can make some extrapolations based on the content of a site, but — for reasons I will get to shortly — the idea of presenting an “average blogger” at all can rub some the wrong way, and to do so without appealing to a more fitting method (namely, some form of survey or interview), and the seeming assumption that self-revelation introduced little bias, was a bit odd.
Mind you, the HICSS paper more clearly defines the effort as an attempt to describe the formal elements of a weblog, which is a process I would be far more sanguine about. And certainly “how much information about the blog author appears in the blogs” seems like something entirely worthwhile to measure.
I see a couple of issues with the sample as well. The first has to do with the potential for tautology: you risk the danger of sampling a universe of blogs that you then in turn define by that sample. It’s the equivalent of saying “I want to describe common characteristics of Gen Xers. Bob, George, and Frank claim to be Gen Xers. They are all men. Therefore, all GenXers are male.” That is an extreme exaggeration, of course, and especially so if “Gen X” is more than simply a self-ascription. In fact, what you did was take blogs that pinged a particular ping-server, then cut out the ones that didn’t fit your provisional definition of blogs, and then extrapolated from that content to a genre description. I think this raises a methodological problem: how do you sample an “emergent” genre without first defining its limits, and how do you then avoid participating in that “emergence.”.
*The average blog*
Elijah dismisses (if that is the correct word) my paper on Park as explicitly sociological, which, given the scholar being called upon, is perfectly fair. I didn’t consider my presentation to be a research paper, it was designed — although I don’t think it succeeded in this — to provide an orienting point for the panelists that followed. Park was adept at using both ethnographic and what we would now call more traditional quantitative social scientific approaches to measuring human environments. I am curious how Elijah would characterize his group’s work if it is not sociological. Rhetorical?
Indeed, his project to find the “average blog” calls to mind the beginning of statistical sociology, and Adolphe Quételet’s effort to find the “average man,” from which abnormalities could be discerned. It’s easy to forget that this was a significant step in the understanding of social behavior. All the same, it was ultimately a pretty wrong-headed approach to better understanding man. The average human is just over 50% female. While this statistic is of some very small descriptive benefit, it would be far more interesting to know how the average female differs from the average male, or how that distribution differs in the US and, say, India. (Likewise, comments like “the median blogger is a 14-year-old boy,” while they may be true enough, don’t tell us much.)
Moreover, I would suspect that these averages are changing fairly rapidly with both the evolution and diffusion of blogging. Therefore, a “snapshot” may be of limited use. The whole reason for doing descriptive work is so that analysis may be built on top of it, but if that description is highly mutable, it makes it difficult to build from it.
But far more important, and I think a significant epistemic break between Liz’s position and Elijah’s, is that the average blog sucks.
That is to say, the social effect is unlikely tied to blogging as whole technology or genre. I would argue that the most exciting unit of analysis is a community or self-identified grouping of weblogs. But even if you don’t accept this, I think it is fair to say that the greatest social impact of blogging so far can be traced to a very small sliver of the extant blogs. Moreover, the _kinds_ of effects these different subsets of blogs have are important.
I think the HICSS paper gets at this to a certain extent. The function of the blog matters. But even then, it’s hard to say that “blogs” as a whole are anything so alike that the average makes any sense. Fark and Scoffing at Gravity both hit me as being “blogs” and fit many of the common criteria, but taking the average of these (and note that the latter type is far more prevalent — how many “farks” are you going to get in a small sample of the blogging world, and yet the impact of such a site is several magnitudes greater than that of the “average” blog) seems to me to be a step in the _wrong_ direction if you ultimately want to be producing some theory-driven, and theory-driving, explanatory research. Sampling relies on the assumption that the universe is reasonably close to a normal distribution, and for the most important characteristics of blogs – the differences that most make a difference – this is not the case.
So, the “average blog” is an attempt to answer Liz’s call for a more empirical, thoughtful description of blogs and blogging. And I have to agree with her that Elijah’s work, while certainly valuable, misses some of the texture of text-in-process. That is to say, it lacks what all content analysis _must_ lack in order to remain valid and reliable, an understanding of the latent social construction that occurs in writing and reading a blog. But from my perspective, it isn’t that the answer is in some way “wrong.” In fact, I in many ways prefer the empirical, quantitative description that you provide to ethnographic observations that suffer from a lack of generalizability. I just don’t know that the _question_ gets us very far. I’m much more interested in how social structure changes, and how human individuals change, when they engage with one another using this technology. I know both Liz and Elijah are deeply interested in this, too, and I suspect that they each hope descriptions and definitions of blogs will provide a springboard for a further understanding of these social dynamics, but I remain skeptical of how such research can be employed analytically.
*Peer review sucks, too*
As a final note, Liz has good things to say about blogged review, and Elijah has good things to say about traditional peer review. I find little good to say about either. When I post academic work to my blog, very few of the comments I get back are of the quality and depth that I get in my best academic referees’ remarks. Likewise, the vast majority of referees that review my work in more traditional academic settings so narrowly define what they consider worthy (and this is especially true of research in blogging which is both necessarily interdisciplinary and tainted by novelty), that the comments invariably may be reduced to “that’s not how I would do that, so you shouldn’t either.” I wouldn’t look at blogs in the ways that Elijah or Liz do, but I also would, as a peer reviewer, recognize the merit in their approaches. I wish more reviewers would.
I think both blog-review (i.e., early-term or pre-publication review) and traditional peer review have their place. I also think that the informal nature of the former may be able to help rectify some of the sins of the latter. But having put in more than my requisite daily blogging time, I’ll leave such a discussion for a later date.