Note: there has been a great response to this. Lots of people are interested in either building something or suggesting collaborations. I’m trying to organize the various ideas and will post follow-ups here. In the meantime, if you’re on github, I’ve thrown up a stub with some very preliminary design notes, please contribute! https://github.com/paultopia/scholaw/
Elsevier just bought SSRN. Here’s why you should be upset, and what we can do about it.
For those of you who aren’t familiar with the academic publishing racket, SSRN is (was?) a huge repository of open-access preprint papers, used all the time by law professors such as myself, but also by people in many other disciplines. As far as I can discern, their business model has been:
- Free for individuals to upload and download papers,
- Sell subscriptions to “e-journals” to universities, where those “e-journals” are just lightly (and often badly) edited e-mail lists of the latest uploads in a given subject area. The idea being, of course, that individual faculty at those schools then have a good way to keep up with the latest research in an area.
Elsevier is the most notorious of the giant for-profit academic publishing firms. They charge truly astonishing subscription costs for academic journals, even though they aren’t obliged to pay authors, referees, etc. For example, one of the more egregious of their yearly subscription rates is the Journal of Nuclear Materials, which will cost libraries $7,442.14 for an electronic subscription, or $11,164.00 for a print subscription. (Full disclosure: I cherry-picked that one for impact.) No, those numbers are not typos. If your institution doesn’t have a subscription, a single article download will cost you $40.
Of course, as an author, you can always choose to make your paper open access. The fee for doing so? $3,500. Apparently some authors actually pay these fees (or, more to the point, have them paid by grant agencies), because the professional validation of peer review is so important for academic careers that it’s worth it to get the stamp of a professional journal on one’s work even if it means that money gets flushed down the toilet for the privilege of getting to post one’s own work on the internet. (It’s basically a three-and-a-half grand web hosting fee — and for the price you could get almost sixty years of a Digital Ocean VPS and serve 20 gigs worth of free papers.)
Unsurprisingly, these business practices have led to a lot of opposition. Last year, the entire Dutch University system was on the verge of boycotting Elsevier over their shameless rentier publishing model, and managed to coerce a little bit of open access out of them (but other countries don’t seem to be able to get it together to follow suit). There is an ongoing attempt to create a worldwide boycott by individual academics. Last year, the entire editorial board of the journal Lingua resigned over exorbitant open access fees. Elsevier of course files a healthy load of copyright lawsuits. It’s tried to co-opt Wikipedia to become a marketing platform for their $40 pdfs. The stories just go on and on.
How did Elsevier get to a position to be charging these prices? Are they just providing a better service? Is the market working? Well, not as far as can be discerned. Rather, they seem to be very good at acquiring smaller publishers; their massive profits are driven by an uncompetitive marketplace.
And now they’ve turned their acquisitive appetites onto the tools that researchers use to communicate with one another mostly for free. It started with Mendeley, formerly a fine open tool for citation management and paper sharing. And now SSRN.
So what’s going to happen to SSRN? Well, it’s unclear, but note the careful weasel words in the official announcement (emphasis mine):
The majority of SSRN’s content consists of working papers, the versions of which Elsevier has always been open to sharing, and they’ve done a lot of work to clarify which versions of content can be shared. Both existing and future SSRN content will be largely unaffected and, like Mendeley, we’ll help researchers share post-submission versions of their work responsibly
That language gives Elsevier a lot of room to enforce its own conception of what “responsible” sharing means, or, in short, to adopt aggressive interpretations of its own contract and copyright rights and then implement those interpretations by fiat on the massive platform that it now owns, while leveraging network externalities to make it more difficult for academics to leave that platform for somewhere less captured.
So what to do?
This could be an opportunity. SSRN was never all that great — it lacks search facilities, relies on human “editors” to curate content, and has a mysterious pricing model. Who needs it?
Instead, the community can come together to create something better. Here’s my rough, preliminary, sketch about what that better thing would look like.
- Organized as a nonprofit or trust with legal obligations to offer its services for free and remain unowned by any for-profit entity (to protect it from future Elsevier buyouts).
- Centered around a web application that allows researchers to upload their papers, and tag them by subject matter, which are then searchable by (a) subject matter tag, and (b) full-text. Anyone can read and search, with no account required or any other restrictions whatsoever.
- To maintain quality and replicate the effect of SSRN’s curation, but without the human labor, uploading papers initially open only to faculty members, postdocs, VAPs, and others with terminal degrees and active university affiliations, but expanding over time to independent scholars, graduate students, etc.
- Automatic tag-based e-mails, replicating the effect of SSRN’s e-journals, but again without the human labor.
I envision the tagging system as the the key feature here. As far as I can tell, the only value-added that SSRN provides is that they’ve managed to get some humans to look at papers and tell whether they really belong in the “Cyberspace Law eJournal” or the “Democratization: Building States & Democratic Processes eJournal” (just to pick two of the SSRN e-mails in my inbox at the moment). But human intervention is not necessary to serve this function, under the following conditions:
- The people allowed to post are filtered at the front end.
- The set of subject matter tags is initially limited in order to avoid thousands of meaningless or redundant tags (perhaps a system of primary tags created by the system and secondary tags created by posters, with the possibility for the secondary tags to be promoted to primary tags with sufficient usage).
- Readers are given an option to provide feedback on subject matter tags (e.g., by downvote, where a sufficient number of unique downvotes removes the tag from a paper).
- Perhaps ultimately leveraging text-mining machine learning techniques, once the quantity of papers is sufficiently large, to automatically correct or supply subject matter tags.
This system would be easy to create. The main challenge would be marketing, convincing people to take their SSRN papers and put them on this alternative service. (An automatic transfer tool would be good for that, though I imagine Elsevier would quickly block it from SSRN servers and/or start filing lawsuits.)
I can’t do it all on my own, mainly for lack of time, but I’d love to talk to others who are interested in doing it collaboratively. Who wants to help me put SSRN down before Elsevier gets its hands on it?
There’s been a quite active discussion on Twitter, Facebook, etc. — see especially this Twitter conversation. A number of people have pointed out various disciplinary open-access repositories, including:
- the grandparent of them all, arXiv in physics, math, CS,
- bioRxiv in biology,
- Philpapers in philosophy,
- A nascent attempt in sociology, SocarXiv (not yet running),
- ADS in astrophysics, as well as INSPIRE (formerly SPIRES) in high-energy physics,
- MLA Commons CORE in humanities, and
- RePEc and MPRA in economics.
But, as far as I can tell, we don’t have anything that has the cross-disciplinary reach of SSRN, and also nothing that covers my own disciplines of law and political science.
So is the disciplinary siloed approach the way to go? (In which case, fellow lawyers and political scientists, let’s create something.) Or is there distinctive value in the cross-disciplinary hub?