Today, SC saw a mention of this site, which purports to organize blogs into a biological taxonomy-type system based on the number of links to them. Naturally, SC's immediate concern was: "how high do I rank?". Being quite humble, the SC research staff started out at the very bottom, with "insignificant microbes", and was relieved not to turn up there. In fact, they had to go almost halfway up the hierarchy to "flippery fish", in order to find the SC listing. If this sort of ranking can be believed, we're #3230 out of some 7600 blogs. The top 50% sure beats SC's finishing position in the last half-dozen ESPN NCAA Tournament Challenges, and so as with all data that makes SC feel good, it is hereby accepted as reliable.
As for the 800-pound gorilla of linguistics blogs, they are in fact a tiny, lightweight bird, at #1307 ([which would make them a kingfisher to your salmon, pal -- ed.]).
Of course, what this is really measuring is the propensity of other bloggers to find your comments interesting/inflammatory enough to be worth linking to. If bloggers never did anything but write about each other, the whole thing would become a self-referential morass, and there would be no original commentary. This isn't so much about popularity as influence, though, and it seems to SC that the extent to which other people feel a need to take up your ideas is a good measure of how influential you are. Sometimes the connections are very much random -- when SC borrows from Andrew Sullivan, it's usually not to respond to his arguments, but rather to make some cheap linguistic point. As measurements go, though, this one strikes SC as being a pretty good one for what it's trying to capture. What would make it even more useful as a measure of influence would be having some way to identify the percentage of the links that indicated agreement/disagreement/bizarre SC-style comments. That qualifies as an ugly NLP task which nobody is likely to pay SC to develop a solution for.
Readers who find this sort of thing interesting beyond fanning the flames of SC narcissism might like to look up the DARPA program "Evidence Extraction and Link Discovery", which looked into some of these issues in a more systematic way (although with serious concerns about privacy, which led to the termination of the office that ran it). The original EELD pages have been largely purged from the DARPA website; some of them are archived here.
Yo, if there's an 800-pound gorilla of linguistics blogs, it's Languagehat, who weighs in as a Large Mammal at #256 overall. Grawrr!
Posted by: language hat | February 12, 2004 at 01:21 PM
I must be a virus or prion, since I don't even show up in the list :)
On a separate note, I also take offense at the title "insignificant microbes" since all life as we know it relies on these "insignificant" critters. But oh well ...
Posted by: Rhosgobel/Radagast | February 13, 2004 at 07:37 PM