Open Access Debate.
I’ve written here about the Open Access (OA) debate many times. On Tuesday evening, at Pub Style Science, I’ll be participating in a friendly debate about the merits and shortcomings of OA scientific publishing. A quick overview:
Open Access – A system by which papers are published online and freely available to anyone with an internet connection. Authors or their institutions generally pay anywhere from $1,000 to $5,000 to publish. The good OA journals (like those run by the Public Library of Science or Biomed Central), conduct rigorous peer review by qualified editors and reviewers.
Traditional Publishing – A system by which authors and institutions generally pay nothing (but some journals may have page charges) to publish. The good traditional journals conduct rigorous peer review by qualified editors and reviewers. Libraries pay for access to the journals, and those without library access can buy individual articles for anywhere from $20-50, usually. Individuals may also purchase personal subscriptions, though they are often very pricey.
Fundamentally, the only differ only in the “who pays” aspect. But many other issues confound the debate. Many OA journals are deliberately non-exclusive. The journal PLoS One, for example, accepts every paper which passes peer review, and reviewers are explicitly told not to judge papers on novelty or impact. Only on the soundness of methods. To my knowledge, no traditional journal does this. This results in the inability to use the journal’s reputation as a proxy for the impact (and perhaps quality) of the science published in it.
Now, for some, this is a feature: science should be judged on its own merits. Read the paper! Judge its quality yourself! For some, this is a bug: I can’t possibly evaluate every paper, especially those not in my direct field. Using journal-reputation as a proxy allows me a higher probability that the papers I find are important, interesting, and valid. There are flaws, of course. Bad science gets published in good journals fairly frequently. But as a general guideline, well-known prestigious journals usually publish science that is correct, interesting, and important.
Other confounding factors: a huge number of bogus OA journals have sprung up, allowing the patina of legitimacy to be appended to bad science, or pseudo-science. It’s true that there are bogus traditional journals as well, but the fact that you must pay for access means that they are less likely to be exposed to large numbers of the public. Paying for publication means that researchers must often use grant money to pay for publication that otherwise would go to research, further stretching limited budgets. Though this is also true of traditional journals with page-charges.
The revenue model is troubling too, for critics of OA. Because journals make more money the more they publish, and are generally unrewarded for how widely their published articles are distributed, they have incentive to publish as much as possible. It’s easy to see how this can result in pressure to publish more, and scrutinize less. As long as the journal is run by the original True Believers, it may maintain integrity. But as time passes, and the original founders retire and are replaced by professional administrators, will that continue?
Another argument that is made is that OA allows for the publishing of small results that traditional journals will not publish because they’re not impactful enough. And that’s true, it does. But, it’s also easy to see how the pay-for-publication model will actually suppress publication of small results. Labs with limited budgets may well be unwilling to publish small or negative results because they cost money. They may well want to save their publication fees for what they decide are “important” papers. After all, they have careers. However they’re published, impactful articles are always going to be important and desirable. A traditional low-impact journal may be precisely what is needed to publish that negative result.
I will be staking my standard territory, for which I’ve been vilified: traditional publishing is not evil, both models have flaws, and there is room for two models. Furthermore, I’ll be reiterating the basic philosophy of my field of study: the publishing system is a massive complex system. When OA proponents tell you what will happen when OA is universally adopted and traditional publishing is dead and buried, they’re making it up. They can’t know. It is a non-linear, hybrid, time-dependent, memory-dependent dynamic system. And we can prove that we can’t predict how those will behave. With a good model, we might be able to gain basic insights, but no one has built the model yet. And it would be a vast and baffling undertaking.