Tuesday, May 24, 2005


Here is the Pew Internet and American Life Project's report called "Buzz, Blogs and Beyond" [only click the link if you want to download the pdf].

What a strange report. The authors want to empirically test claims they say are being made, loudly and often, for the significant influence of blogs on Politics and Big Media in America. Their set-up trots out all the chesnuts about how popular blogs are (number of users in the US, number of readers, etc.), and although they nowhere evince out and out suspicion, their impulse to *test* these claims seems to me reactionary from the outset. But let's leave that aside for now.

Their metric and their method, it turns out, is right there in the title: it's all about "buzz." It will come as no surprise then that the study was led by Dr. Michael Cornfield in collaboration with a company called Buzzmetrics. Put tersely, "buzz" is "a lot of simultaneous talk" (p. 2). But the most we get about it is the still pretty terse: "Buzz is the sound heard in public when a lot of people are talking about the same thing at the same time" (p. 3). Then they go on to briefly list some of things "buzz" can do: affect sales, alter behaviour and perceptions, embolden or embarrass, and "move issues up, down and across institutional agendas."

With "buzz" as their metric AND the primary phenomenon being measured, their study has some circularity problems. If "buzz" is defined at all, it's defined tautologically: "buzz is the sound heard when...". Yes, but, how do people hear it, how does it get amplified or distorted or disrupted? What are its mechanisms, its modes of circulation? But, ok, let's leave that aside too.

What do they find?

1. "Blogger power is circumstantial: dependent on the sorts of information available, and contingent on the behavior of other public voices."
2. "Political bloggers were buzz followers as much as buzz makers."
3. "Bloggers may have been positioned in the fall of 2004 [when they conducted their study] as a guide for the mainstream media to the rest of the internet."
4. "The key contribution by the bloggers in the "Rathergate" scandal consisted of providing forums accessible to all internet users in which facsimilies of the memos could be examined and discussed."

Subsequently, the tone of the paper is measured, but it's as if they're having to try...very...hard to stay neutral. When they cede some measure of influence to blogs, they do it begrudgingly. You can whiff it in those findings from the "executive summary" above. And I'm not the only one who has noted it. So we get this article, whose key "take-away" from the Pew study is telegraphed in the title: "Blogs May Not Be as Influential as Some Think".

Aside from the questions we might want to ask about this need to measure, test and confirm/deny the vaunted "influence" of blogs (which I've written about in "A Welcome for Blogs", forthcoming 2006, Continuum), I have two basic reactions:

1. The concept of "buzz" seems to belie (to bury rather than represent) almost everything interesting about the way that blogs operate.

2. The study admits of a very narrow scope for "influence:" to be influential on their terms a blog or set of blogs must be seen to (measurably, quantifiably) set the agenda for popular discussion. That is: lead discussion, be originary, avoid being derivative. This seems to me to impose a hegemonic model of influence (in which there can only be leaders and followers) on a form of public discourse which, in some quarters, is doing a lot to challenge that very model. So, while we hear a kind of tactful schadenfreude in the finding that "Political bloggers were buzz followers as much as buzz makers," if the study were more attentive to the characteristic rhetorics of blogging, it might recognise in that finding not only a unique form of influence (to "follow" is to relay), but also the ways that bloggers are part of a larger process of changing how we might define, recognise and assess influence, in the first place.
web statistics