I recently reread Paul Ginsparg’s article First Steps towards Electronic Research Communication, about the establishment of the physics e-print archive. I reread it partly because I’ve been thinking about institutional repositories lately, & partly because it’s 10 years old recently (the article is dated 19 April 1995).
In it, Ginsparg writes:
This “e-print archive” began as an experimental means of circumventing recognized inadequacies of research journals…
This system provides a paradigm for recent changes in worldwide, discipline-wide scientific information exchange, and a model for electronic transmission of research and other information…
And so I just have to ask: ten years later, why are we still having this conversation? All of these “recognized inadequacies of research journals” still exist: the serials crisis still exists, times to publication are still absurdly long, the list goes on. Plus, the web makes dissemination of publications trivial. So I ask myself, why has every field not adopted an e-print archive?
Even allowing for differences in technological savviness between fields… I mean, sure, it would be the high-energy physicists who developed the first e-print archive. They always seem to be on the cutting edge of information dissemination. (And why is that, exactly? Someone in information science should do an ethnography of high-energy physicists.) But look, 10 years should be enough to allow even the most technologically backward field to catch up, if the will was there.
So clearly the will is not there. Not even in information science, amazingly; we don’t have an e-print archive either. Why not? Well, for one thing, HEP, CS, and other fields move so fast (no pun intended about HEP, really) that by the time an article is published through traditional journal publishing, it’s ancient history. The journal as archival medium, rather than as vehicle for information dissemination, blah blah. This is less the case in IS: this field moves fast but not that fast.
So why don’t more fields have e-print archives? I can think of only 1 reason: peer review. No one has yet come up with a good method for implementing a peer review system in an institutional repository. Oh sure, Harnad discussed this 15 years ago, but no one has actually implemented it.
So ok, we like our peer review. But just because journals aren’t a field’s primary publication medium doesn’t mean that peer review goes out the window. CS has very effectively moved over to conference pubs being the primary publication medium, & some CS conferences have lower acceptance rates than many journals. How did CS make this transition? I don’t know; someone should do an ethnography of computer scientists’ publication behaviors while they’re at it.
Somewhere on the continuum between journal and e-print archive (more towards the journal end, I believe) lies the open-access journal. The recent news of PLoS Biology‘s impact factor, and the fact that it’s so high, seems to me to be definitive proof that open-access journals can play with the big boys. Seems to me that there are 2 criteria for evaluating journals, one based in scholarship per se and one based in the social processes of scholarship (or, I suppose, one idealistic & one cynical): their effectiveness as vehicles for disseminating information, and as generators of reputation & status for its contributors. These are of course closely interrelated. Seems to me that given PLoS Bio‘s IF of 13.9, OA journals meet both critera, QED.
IRs fulfill 2 of the 3 functions of journals already: dissemination & archiving. But this is totally a case of “the operation was a success but the patient died”: there are a bunch of criteria that we can use to evaluate the success of an IR in scholarly communication (of any form of scholarly communication, really), but really only one that matters. If someone could figure out how to make an IR fulfill the function of generating status for its contributors, then it would fulfill all 3 functions of a journal.