Much ado about rather little

H.H. Wills Physics Laboratory, University of Bristol To people outside science, and to many scientists, open access seems so obviously good that to object seems unreasonable. But I have doubts. First, we should get it in perspective. The whole open-access discussion is a fuss about an almost non-existent (and as far as physics and mathematics are concerned completely non-existent) issue. We are invited to imagine a member of the public wanting to read a scientifi c paper and without access to a library. There might be a few such people and we would surely want to help them. But almost every paper will be freely available already, either on a repository such as the arXiv or on the author’s home page. And for papers that are not available, I cannot imagine that any scientist, when asked by an interested reader to send a copy of a scientifi c paper, would refuse to do so. Such openness has long been common practice: nowadays we email our papers as pdf fi les; in previous decades we sent photocopies by post. That said, in principle I have no fundamental objection to the authorpays model. Although I was initially sceptical about online-only, openaccess publications, such as the New Journal of Physics, their success, and my own experience, have dispelled these doubts. But if open access is to become the norm, there are two serious points that must be addressed. The fi rst concerns scientists with little or no research funding – retired people like me, and scientists from poor countries. Several publishers have reassured me that in such cases publication charges would be quietly waived. ‘Quietly’ is probably wise – that is, not an explicit policy but a practice. Such waivers are essential: without them, a large class of active scientists will be excluded from publishing. Open access for readers would lead, by the law of unintended consequences, to restricted access for many scientists – a cure far worse than the (non-existent) disease. Now the second point. UK scientists applying for grants from the research councils will be forbidden from including open-access publication costs in their applications. Instead, these costs will come from university budgets. The superfi cially reasonable argument for this is that the costs will replace money currently spent by libraries on journal subscriptions. Here I see three dangers: (a) scientists will have to apply, and probably compete, for these costs, leading to a new level of bureaucracy and unacceptable delay; (b) for a prolifi c researcher, publishing say 20 papers a year (not unusual in theoretical physics), the cost would run into tens of thousands of pounds, raising the possibility of a university questioning the need to publish so many expensive papers; (c) worst of all (and as others have pointed out: Letters, London Review of Books, 24 January 2013), universities might start questioning the journals chosen by authors, and pressing them to publish in those with high impact factors. All this, to solve a problem that barely exists! If open access comes to dominate, I would hope that other funding agencies (the Royal Society, Leverhulme, Wellcome, etc.) will refuse to follow the research councils; they should allow their researchers to claim the costs of open-access publications. This would send the powerful message that RCUK’s decision to shift responsibility onto universities is not the only way. And when RCUKfunded university researchers, mired in university bureaucracy, see their otherwise-funded colleagues publishing effortlessly, perhaps they will rise up and demand a return to sanity.