Guest post written by Philippe Desjardins-Proulx
arXiv is arguably the first great success of open access in science. Created by Paul Ginsparg 21 years ago as a repository of preprints, arXiv hosts more than 700 000 preprints from physics, mathematics, computer science, quantitative biology, and other fields. arXiv is not a scientific journal in the traditional sense, it is a place where scientists submit their preprints, almost always before submitting the paper for peer-review. arXiv has moderators who can move preprints from one section to another, but no review is made. In some areas of mathematics and physics (notably high-particles physics), almost all papers are submitted to arXiv before being subjected to peer-review. I want to emphasize this: arXiv is not a substitute to peer-reviewed journals, it is a way for researchers to share preprints before the papers are submitted to peer-review. The main idea was to improve the scientific process, not to avoid it. Instead of sharing preprints with a select few colleagues, scientists share their product with the entire community. The preprints can then be shared and discussed openly as soon as they are finished and submitted to arXiv.
Surprisingly, while high impacts journals like Nature and PNAS and most quantitative sciences have embraced arXiv, many ecology journals refuse to consider papers submitted to arXiv. According to SHERPA/RoMEO, all major scholarly publishing businesses accept that authors archive pre-prints (Wiley-Blackwell, Springer, Elsevier, etc.). Yet, journals like Ecology Letters, Global Ecology and Biogeography, and the journals from the Ecological Society of America consider arXiv a form of “prior-publication”, despite the fact that these papers have not being reviewed and are fundamentally no different from exchanging manuscripts with colleagues or presenting results at a conference. I believe that a strong case for arXiv can (and should) be made, and that organizations like the Ecological Society of America should be pressured to adopt more progressive policies. arXiv is part of a vigorous scientific process and we have no reason to be bullied into a closed system while other fields benefit from an open process.
arXiv is not a new adventure and is now the norm in mathematics, high particles physics, condensed matter physics, and many other fields. We thus have a good idea of its benefits.
arXiv improves science and makes it faster. The traditional model of science was not significantly affected by the development of the internet. Of course, papers are now submitted electronically, but the process remains long and most often done behind closed doors. arXiv is part of a broader perspective on the scientific process. We often reduce the process to the peer-reviews requested by journals, but prepublication reviews by a small network of colleagues is also a significant filter and an important aspect of a vigorous peer-review process. arXiv is simply a way to extend this network of colleagues to the entire scientific community. It ensures that science is not constrained by these small networks of scientists exchanging ideas. Ginsparg made arXiv in part for democratic reasons: he wanted everyone from graduate students in small universities to Princeton professors to have access to the most recent scientific ideas.
arXiv avoids some annoying problems with closed-doors reviewing system. We’ve all heard story of this kind: A submits a paper, B gives a bad review and submits a similar idea. A common argument against arXiv is that it allows other researchers to publish your ideas before you do. Actually it’s the complete opposite. arXiv is a place to claim intellectual precedence and publicize your own work so other scientists can see what is currently being done. It’s also a much fairer system to establish priority, as some journals will take more time than others to publish articles.
Why are so many ecology journals hostile to arXiv? Honestly, I don’t know. It’s puzzling. All high-impact journals (Nature, Science, PNAS) tolerate arXiv, with Nature and PNAS being quite enthusiastic about it. It seems like a win-win situation. arXiv is no substitute for scientific publications and journals can still sell the finished product, yet the paper is more visible, known to many specialists even before being published, and priority is established. I am not aware of any evidence that journals in high particles physics are in any kind of trouble because of arXiv.
The common argument goes that arXiv is a form of prior publication, or that arXiv is detrimental to science because of its permissive nature. These two arguments are patronizing. Everyone knows that arXiv papers are meant to be submitted to peer-review journals. They have no more, but no less, credibility that an open discussion in a lab or a talk. Yes, the permissive nature of arXiv means that some crackpots are submitting crazy theories (although it’s surprisingly rare and the moderators push some in General Mathematics/General Physics). arXiv papers can be cited, just like personal communications are something quoted, but it’s the job of scientific journals to give the paper legitimacy through the peer-review process. Again: arXiv is not a way to avoid review, on the contrary, it’s a way to invite reviews and make the process of science more open, more vigorous. It’s based on the philosophy that peer-reviews are simply one part of the scientific process, along with prepublication and postpublication reviews. If Nature, PNAS, Science, mathematics and physics journals are not threatened by arXiv, why is it different for the ESA and Ecology Letters? Ecologists have less rights? Are not smart enough to understand the role of arXiv? The notion that some copyright issues would prevent them to accept arXiv papers begs the question: why is it not a problem for Nature, Springer, Wiley-Blackwell et al.?
Not only is arXiv not the enemy of peer review, it is increasingly seen as an important part of a vigorous review of scientific evidences. Michael Hochberg concluded his great piece on the importance of good peer reviews with three themes to improve science. His first point was the need for…
Online science, such as ResearchGate or arXiv, where manuscripts can be continuously updated based on readers’ comments. In both instances as well as others the verification of the process of scientific improvement needs to be monitored.
So what can be done now? First, we can put pressure on journals refusing arXiv. Boycotting them is even better. In an increasingly open scientific world, the notion that a few journals can dictate what we do with our preprints is hard to swallow. Secondly, of course, ecologists can submit to arXiv or similar services. arXiv has currently few sections in quantitative biology but new ones could be created, especially since the “Quantitative Biology” section is currently growing faster than other sections. The current section for ecology is “Quantitative Biology > Populations and Evolution”, and some sections for theoretical ecology, community/ecosystem ecology, or quantitative phylogenetics would be welcomed. Also, PeerJ will soon offer a service similar to arXiv aimed at life scientists, and F1000Research is essentially arXiv + post-reviews.
The apparent disconnect between scientists and publishers raises many questions, especially for journals supported by non-profit organizations devoted to the advancement of science. Hubbell (1) made a strong case that developing unified theories to understand biodiversity should be a priority for mankind, as the destruction of ecosystems threatens our well-beings (and, worst, threatens our objects of studies!). Our science ultimately depends on publishers and their policies. We should make sure that the foundations on which our science is based are solid. If some societies and publishers slow down the progress of science with conservative policies, we should turn to more progressive institutions. There are plenty of good arXiv-friendly journals. “No community that has adopted arXiv usage has renounced it” and I hope that ecologists will soon adopt this extraordinary (and modern) way to do science. Recent discussions on the topic are certainly encouraging.
I thank Timothée Poisot for the opportunity to write this post, Ethan White for a wonderful Twitter discussion that ultimately lead to this post, and Hedvig Nenzén for comments on an earlier version.