Sci-Hub, the massive online repository of 50 million academic papers, chapters and books, was set up by a postgraduate student from Kazakhstan in 2011. Alexandra Elbakyan has said she her goal was to provide free access to research for people like herself for whom access was not provided by her institution and too expensive to pay for each article she needed for her PhD. She used login details of others (she claims provided willingly) to gain access to pay-walled research databases and add documents to the site, which, despite the name, includes non-scientific literature as well. Depending on your perspective, she’s either a Robin Hood providing access to the poor or a copyright pirate on a massive scale. Certainly what she has done is illegal — and she is being sued by Elsevier — but Sci-Hub’s existence and the widespread use that’s being made of it by people around the world says much about the state of scholarly communication and open access to knowledge.

If you don’t know what Sci-Hub is, think of a Google-style search box, where you type in an article title or other identifying information and, presto, the research you want appears in PDF format. No pesky institutional usernames/passwords, multiple-clicks inside learning management systems or searching different databases. It’s all there in one place. Try out a couple of article titles and you’ll likely find the work. If you get an exact match (e.g. using a DOI, a digital object identifier), it will appear immediately without even a the inconvenience of a search results page.

John Bohannon has published a fascinating piece in Science analysing data he obtained from Elbakyan herself. For one thing, he says, Sci-Hub is effectively “the world’s de facto open-access research library.” What he really means is “free access,” since open access implies a licence that allows reuse, but the point is well made. He obtained data from Elbakyan covering 6-months of use of the site: a staggering 28 million hits were recorded in that period. There is location information in the dataset too, which, while not necessarily 100% reliable, shows use not only across the globe but, interestingly, clustered around locations where people are likely to have access through an institutional subscription to the legitimate source. This suggests Sci-Hub is not only used by people who don’t have access and can’t afford to pay but also by those who just find it more convenient than conventional means. Check out the data in Bohannon’s piece, which has an interactive map (made possible by open software and data I should point out). There has been just under a couple of hundred accesses by people in the Dunedin area. Someone in Timaru has been reading Museum perceptions and productions: American migrations of a Maori hei-tiki. I didn’t even know there was a place called Mayfield on the Canterbury plains but there are 33 hits recorded for that location and at least one person is apparently interested in engineering and minimum design loads for buildings and other structures. This could be someone studying or teaching engineering, who might have legal access to such research, but equally it could be a private citizen who has a Cantabrian’s interest in things affected by seismic activity.

Will Sci-Hub be the ‘Napster moment’ for academic publishing, a development that challenges the established model to the extent that it is forced to change as the music and other industries have? Some long-time open access commentators hope it will, such as Peter Murray-Rust, who has written a three-part opinion piece on why Sci-Hub matters (part 1 here). The law is on the side of rights holders like Elsevier but then it was on the side of the music industry, which may have beaten Napster in court but was nevertheless forced to change its model by consumer demand.