An Academic Publishers’ GO FAIR Implementation Network (APIN)

Presented here is a proposal for the academic publishing industry to get actively involved in the formulation of protocols and standards that make published scientific research material machine-readable in order to facilitate data to be findable, accessible, interoperable, and re-usable (FAIR). Given the importance of traditional journal publications in scholarly communication worldwide, active involvement of academic publishers in advancing the more routine creation and reuse of FAIR data is highly desired. 1. Academic publishing since the emergence of the Web The proposals in this article must be seen in the context of developments in academic publishing and further trends we can foresee. The scholarly publishing environment has changed dramatically in the last 25 years. The change was precipitated by the—quite sudden—emergence of the World Wide Web (commonly just called ‘the Web’), the common system for navigating the Internet. The idea that the Web would make the postal service irrelevant for sharing scientific information was suddenly not whimsical anymore, even though in the early days, it seemed a bit of an exaggeration. All information, and therefore also scholarly, scientific information, could be distributed and shared electronically. Its reach was henceforth only limited to access to the Internet itself. And it was not just the potential of worldwide reach that was made possible, it was also the high speed—well-nigh instantaneous—with which information could be delivered. The end of the role of print was in sight. ‘Paper’ took on a different, virtual, meaning, albeit that ‘papers’ were, and are, often enough still printed, though increasingly by the recipient of the electronic version and not in a central printer facility and bound in issues. ‘Print’ itself survived as a concept, now virtual, especially in words like ‘preprint’. With printed journal issues disappearing, postal charges, not an insignificant part of subscription costs, and a significant source of profit for at least some publishers, who used to charge more than the actual postage cost, vanished as well. The Internet, and the Web technology riding on it, also made it possible to distribute scholarly material to a vastly larger audience than had been the case with printed issues, as the marginal cost per copy became an irrelevance and disappeared, to all intents and purposes. *Corresponding authors: Jan Velterop. E-mail: velterop@gmail.com. Erik Schultes. E-mail: erik.schultes@go-fair.org. 0167-5265/20/$35.00 © 2020 – IOS Press and the authors. This article is published online with Open Access and distributed under the terms of the Creative Commons Attribution Non-Commercial License (CC BY-NC 4.0). 334 J. Velterop and E. Schultes / An Academic Publishers’ GO FAIR Implementation Network (APIN) The Web made it possible, at least in principle, to distribute the results of scientific research more widely, more equitably. The one impediment that remained – for Academia – was the cost of access. The technology for distribution changed very rapidly, from print on paper to electronic, but the business model of publishers defraying their costs and making money via subscription charges, proved very resilient. The need to cover costs is inevitable (though efficiency gains could possibly reduce those costs), but it was particularly the desire – especially by the larger commercial publishers – to preserve profit margins that made new business models unpopular. A desire among researchers for open access, whereby the reader would not be confronted with paywalls and costs, grew. For a long time, many publishers resisted. A business model that would deliver open access, would also imply that income could not be realised by charging the reader. Consequently, the author should carry all the cost, or some other way should be found to secure income at the input side of the publishing process. Publishing would have to become a business providing services to scientists in their role as authors rather than in their role as readers. Charges would have to be levied for actual services on an article-by-article basis, and not for the availability “justin-case-it-is-needed” to bundles of articles in journals. In such models, a ‘journal’ would become a ‘label’ attached to an article, and not the bundle of articles bound into issues known from traditional publishing. The convenient opaqueness of the subscription model, in which costs could be allocated, made pricing a fine art (a gamble at the beginning of a new journal, and an estimate to ensure gross revenue preservation once a journal became established – juggling factors such as exchange rate fluctuations, growthor attrition rates of the number of subscribers, increase or decrease in submissions and acceptance rates, et cetera). Instead, costs and profit of an open access publishing environment would have to be covered by what became known as ‘Article Processing Charges’ (APCs), which are inherently less suitable to being kept opaque, so the pressure on being transparent was – and still is – increasing. Consequently, the potential for downward pressure on the ‘profit’ part of the equation could well become high. 2. Changing role of publishers There is more: for decades now, authors’ manuscripts have almost universally been prepared electronically, and the publishers’ role was more or less reduced to arranging peer-review for, and some formatting of, the ‘version of record’ of articles. Of course, sometimes there was some copy-editing, too, but the stories of copy-editing being inadequate or even introducing errors that were not in the manuscript that authors submitted, are not exceptional. The example of the German sz-ligature, ß, being used by authors for the Greek beta, β, or this inappropriate substitution not being spotted and corrected by copy-editors, is illustrative. In print, the ß/β issue may not matter much, as the human eye reads what it expects: ßcarotene looks pretty similar to β-carotene on paper. Electronic versions change that, of course. If read by a machine, the Unicode for ß (U+00DF) is quite different to the one for β (U+03B2). It matters, if articles have to be processed and analysed with the assistance of machines, which is increasingly necessary as a result of the overwhelming numbers of articles that are being published (for example the flood of Covid-19 articles which were being published at the time this article was being written [1]). It also means that copyediting, if done well, remains useful, certainly as long as authors aren’t always as diligent as they should be when composing the text of their articles. If it weren’t for these roles, of arranging peer-review and copy-editing, scholarly communication could have easily escaped the scholarly publishing ‘ecosystem’ (or ‘ego-system’, as it is sometimes mockingly called, due to the strong association with researchers’ career and reputation enhancement consequences and effects). 335 J. Velterop and E. Schultes / An Academic Publishers’ GO FAIR Implementation Network (APIN)