The iTunes of science

Interesting take by Chris Surridge on the "long tail" and open access science. Webgeeks talk about the long tail a lot as one unforeseen beneficial consquence of the internet. Graphs of power-law distributions, such as book popularity, can have long tails of numerous low-selling items that in fact outweigh the few bestsellers. Therefore, while traditional bookshops must perforce stock their inventory with bestsellers to get the best return per unit shelfspace, the likes of Amazon can have effectively infinite shelfspace and reap the rewards of serving the long tail (the Long Tail on Wikipedia; the original Wired article and accompanying bookblog by Chris Anderson; lots of talk about it). eBay does the same thing, enabling the teensy-tiny retail activity that traditional markets can't facilitate, and clipping the ticket as it goes. The long tail isn't just about books or second-hand exercycles - it applies to editorial choices in all media, to all branches of retail, to charitable food distribution in the Third World, all over the place. Chris Surridge muses on how it applies to science citations:

I said that the long tail distribution was pervasive. One place where it occurs is in citation data. Just as with record or book sales there are a few papers that are hugely highly cited, a moderate number of papers that are moderately cited and a vast number of papers that are only cited rarely. The shape of this graph is the same whether you look at the whole scientific literature, a specific discipline or a particular journal. Indeed the shape of this graph, being entirely non-normal, is one of the reasons why the Impact Factor of a journal (supposedly the mean citation of papers published therein) has little power to predict the number of citations received by any of its individual papers. Also exactly like a records or books, just because a paper lies in "the long tail" doesn’t mean that the smaller number of scientists who read it have not found that paper vital to their (working) lives.

I’m no fan of re-inventing the wheel. If there are analogies to be made between the dissemination of scientific research and the selling of pop songs we should try and learn from them. Clearly there are differences too. It isn’t vitally important that the latest pop singles should be freely available to anyone in the world with an internet connection, but for scientific research there is. There isn’t a need that a every tune recorded can be sampled and reused (although you could argue that it is desirable), while there is for scientific knowledge. Nevertheless there are many compelling similarities that I hope can be exploited for the good of science starting with a publication that can publish more papers, across more disciplines than any conventional scientific journal and yet treats the blockbuster and the resident of the tail with equal care and respect.

He claims PLOS One is the iTunes of science. How about