Terry Rudolph over at Cosmic Variance:
There has long been a tension between the academic publishing process, which is slow but which is still the method by which we certify research quality, and the ability to instantaneously make one’s research available on a preprint server such as the arxiv, which carries essentially no such certification whatsoever. It is a curious (though purely empirical) observation that the more theoretical and abstract the field the more likely it is that the all-important question of priority – when the research is deemed to have been time-stamped as it were – will be determined by when the paper first appeared on the internet and not when it was first submitted to, or accepted by, a journal. There are no rules about this, it’s simply a matter of community acceptance.
At the high-end of academic publishing, where papers are accepted from extremely diverse scientific communities, prestigious journals need to filter by more than simply the technical quality of the research – they also want high impact papers of such broad and general interest that they will capture attention across ranges of scientific endeavour and often the more general public as well. For this reason it is necessary they exercise considerably more editorial discretion in what they publish.
Topics such as hurdling editors and whether posting one’s paper in preprint form impacts negatively the chances of it being accepted at a high-end journal are therefore grist for the mill of conversation at most conference dinners. In fact the policies at Nature about preprints have evolved considerably over the last 10 years, and officially they now say posting preprints is fine. But is it? And is there more to editorial discretion than the most obvious first hurdle – namely getting the editor to send the paper to referees at all? If you’re a young scientist without experience of publishing in such journals (I am unfortunately only one of the two!) perhaps the following case study will give you some pause for thought.
Last November my co-authors and I bowed to some pressure from colleagues to put our paper, then titledThe quantum state cannot be interpreted statistically, on the arxiv. We had recently already submitted it to Nature because new theorems in the foundations of quantum theory are very rare, and because the quantum state is an object that cuts across physics, chemistry and biology – so it seemed appropriate for a broad readership. Because I had heard stories about the dangers of posting preprints so many times I wrote the editor to verify it really was ok. We were told to go ahead, but not to actively participate in or solicit pre-publication promotion or media coverage; however discussing with our peers, presenting at conferences etc was fine.
Based on the preprint Nature themselves published a somewhat overhyped pop-sci article shortly thereafter; to no avail I asked the journalist concerned to hold off until the status of the paper was known. We tried to stay out of the ensuing fracas – is discussing your paper on blogs a discussion between your peers or public promotion of the work?