The question of inequality and unfairness has come up again, from Seth’s Gatekeepers posts and subsequent conversation, to pointers to Clive Thompson’s A-Listers article in New York magazine, which article discusses the themes from Powerlaws, Weblogs, and Inequality (though without mentioning that essay or noting that the original powerlaw work was done in 2003.)
The most interesting thing I’ve read on the subject was in Doc Searls post:
I’ve always thought the most important thesis in Cluetrain was not the first, but the seventh: Hyperlinks subvert hierarchies.
What I’ve tried to say, in my posts responding to Tristan’s, Scott’s and others making the same point, is nothing more than what David Weinberger said in those three words.I thought I was giving subversion advice in the post that so offended Seth. But maybe I was wrong. Maybe being widely perceived as a high brick in the blogosphere’s pyramid gives my words an unavoidable hauteur — even if I’m busy insisting that all the ‘sphere’s pyramids are just dunes moving across wide open spaces.
[…] I’ll just add that, if ya’ll want to subvert some hierarchies, including the one you see me in now, I’d like to help.
The interesting thing to me here is the tension between two facts: a) Doc is smart and b) that line of thinking is unsupportable, even in theory. The thing he wants to do — subvert the hierarchy of the weblog world as reflected in lists ranked by popularity — is simply impossible to do as a participant.
Part of the problem here is language. Hierarchy has multiple definitions; the sort of hierarchy-subverting that networks do well is routing around or upending nested structures, whether org charts or ontologies. This is the Cluetrain idea that hyperlinks subvert hierarchies.
The list of weblogs ranked by poularity is not a hierarchy in that sense, however. It is instead a ranking by status. The difference is critical, since what’s being measured when we measure links or traffic is not structure but judgment. When I’m not the CEO, I’m not the CEO because there’s an org chart, and I’m not at the top of it. There is an actual structure holding the hierarchy in place; if you want to change the hierarchy, you change the structure.
When I’m not the #1 blogger, however, there are no such structural forces making that so. Ranking systems don’t work that way; they are just lists ordered by some measured characteristic. To say you want to subvert that sort of hierarchy makes little sense, because there are only two sorts of attack: you can say that what’s being measured isn’t important (and if it isn’t, why try to subvert it in the first place?), or you can claim that lists are irrelevant (which is tough if the list is measuring something real and valuable.)
Lists are different from org charts. The way to subvert a list is to opt out; were Doc to stop writing, he would cede his place in the rankings to others. At the other extreme, for him to continue to champion the good over the mediocre, as he sees it, sharpens the very hierarchy he wants to subvert. Huis clos.
The basic truth of such ranking systems is unchanged: for you to win, someone else must lose, because rank is a differential. Furthermore, in this particular system, the larger the blogsphere grows, the greater the inequality will be between being the most- and median-trafficked weblog.
All of that is the same as it was in 2003. The power law is always there, any time anyone wants to worry about it. Why the worrying happens in spasms instead of steadily is one of the mysteries of the weblog world.
The only things that are different in 2006 are the rise of groups and of commercial interests. Of the top 10 Technorati-measured blogs, (Disclosure: I am an advisor to Technorati), all but one of them are either run by more than one poster, or generate revenue from ads or subscriptions. (The exception is PostSecret, whose revenue comes from book sales, not directly from running the site.) Four of the top five and five of the ten are both group and commercial efforts — BoingBoing, Engadget, Kos, Huffington Post, and Gizmodo.
Groups have wider inputs and outputs than individuals — the staff of BoingBoing or Engadget can review more potential material, from a wider range of possibilities, and post more frequently, than can any individual. Indeed, the only two of those ten blogs operating in the classic “Individual Outlet” mode are at #9 and 10 — Michelle Malkin and Glenn Reynolds, respectively.
And blogs with business models create financial incentives to maximize audience size, both because that increases potential subscriber and advertisee pools, but also because a high ranking is attractive to advertisers even outside per capita calculations of dollars per thousand viewers.
(As an aside, there’s a pair of interesting technical questions here: First, how big is the A-list ad-rate premium over pure per-capita calculations? Second, if such a premium exists, is it simply a left-over bias from broadcast media, or does popularity actually create measurable value over mere audience count for the advertiser? Only someone with access to ad rate cards from a large sample could answer those questions, however.)
In his post Shirky’s Law, Hugh Macleod quotes me saying:
Once a power law distribution exists, it can take on a certain amount of homeostasis, the tendency of a system to retain its form even against external pressures. Is the weblog world such a system? Are there people who are as talented or deserving as the current stars, but who are not getting anything like the traffic? Doubtless. Will this problem get worse in the future? Yes.
I still think that analysis is correct. From the perspective of 2003, it’s the future already, and attaining the upper reaches of traffic, for even very committed bloggers, is much harder. That trend will continue. In February of 2009, I expect far more than the Top 10 to be dominated by professional, group efforts. The most popular blogs are no longer quirky or idiosyncratic individual voices; hard work by committed groups beats individuals working in their spare time for generating and keeping an audience.