Martin Paul Eve bio photo

Martin Paul Eve

Professor of Literature, Technology and Publishing at Birkbeck, University of London

Email Books Twitter Github Stackoverflow MLA CORE Institutional Repo Hypothes.is ORCID ID  ORCID iD Wikipedia Pictures for Re-Use

I write to provide feedback in an individual capacity on the Plan S implementation guidelines.

I am extremely supportive of the cOAlition’s goals and Plan S in general. I disagree with those who say that the timeline is too short; many of these actors have not taken the opportunities over the last decade to experiment with open access or new business models and have only begun dialogue under the threat of immediate action. That said, I welcome the recent engagement by the Wellcome Trust and UKRI to speak with Learned Societies and to evaluate routes to their transition to Plan S compliance. Developing alternative revenue streams to support the activities of these bodies is not a small task, but it is crucial for the wellbeing of these disciplines, and for open access to prosper.

There are a few areas where the document could provide greater clarity.

First, it is unclear whether scholarly articles can be compliant by being published in hybrid venues, but without funding coming from a Plan S funder. In other words, at present, the implementation guidelines do not specify whether an article that is CC BY, gold open access, but in a non-compliant journal (e.g. hybrid) would fulfil the terms. By themselves, at present, I would read the implementation guidance to state that such articles are not compliant. However, this is in conflict to the policy of the Wellcome Trust, which has billed its policy as Plan S-compliant. For Wellcome, such an article would be acceptable under its terms, but the Trust itself would not fund such an article’s publication. I would urge the authors to clarify whether their terms pertain to openness of the article or to the uses that funding for publication from members of cOAlition can be put.

Second, it would be helpful if more guidance could be provided on third-party material and licensing. This remains a crucial stumbling block in the humanities disciplines, especially, for instance, art history. While I appreciate that open licensing can be selectively applied to documents, this becomes difficult in the case of sharing and containerisation of third-party materials. For instance, if a PDF file contains third-party material, and a CC BY license is applied to the document, excluding the third-party material, the entire PDF cannot be shared. This may not be entirely obvious from the metadata provided, say, in a repository, which could lead not only to copyright violation but also to documents where the CC BY license has no effect whatsoever, in practical terms. Specific guidance on this – e.g. such as: the compliant version should be the AAM without third-party material – would greatly ease diplomatic relationships with the affected disciplinary communities and provide a route forward.

Third, I am unclear on the legal status of government funders requiring transparency of costs from corporate suppliers. While I am personally deeply opposed to the profit motive in scholarly communications and sickened by the oligopolistic behaviours of large corporate publishers, I am unaware of any other industry that is obliged to provide its costs as opposed to its prices in the commissioning of services, even when funded by the taxpayer. Certainly, I understand that governments feel they should secure the best “value for money” return on public investment. Having a benchmark of the costs vs. overheads can here be helpful in determining whether the value proposition is sound. However, if it is believed that cartel-like price-fixing of standardised rates have emerged that are uncompetitive, then surely a competition inquiry would be the better way to resolve this?

I would also note that transparent costings for publication are very difficult to provide and evaluate as the majority of costs are indirect and substantially affected by unit volume. That is, the cost-per-article can be a highly misleading way of characterising the operation of a publisher. Asking for costs in this way could also perversely incentivize publishers to take on more articles, with fewer staff, certainly driving down costs, but also potentially and problematically motivating poor review standards and quality control. In the bid to present a cost-effective front, there may be unintended and negative behavioural consequences. The recent tender for the ORE platform, for instance, seemed to me to ask a lot for very little and was unrealistic in its requirements against its budget: https://eve.gd/2018/04/01/the-tender-document-for-the-european-commissions-open-access-platform-asks-for-an-awful-lot/

Fourth, the requirement for XML/JATS deposit in the green route is doable, but will require substantial intervention. At present, most researchers are unaware even that their articles are typeset in XML (and do not currently get access to these files from most publishers) and as the current largest green policies – such as Research England’s REF mandate in the UK – have deferred the burden to researchers or mediated deposit by librarians, this is a substantial workflow alteration.

Fifth, while not an ambiguity, I would praise the inclusion of CC BY-SA 4.0 as an acceptable license, which was my recommendation in Eve, Martin Paul, Open Access and the Humanities: Contexts, Controversies and the Future (Cambridge: Cambridge University Press, 2014) http://dx.doi.org/10.1017/CBO9781316161012.

Sixth, I would recommend a broader and more inclusive set of terminologies than “science” throughout. While the European context within which Plan S originated has the humanities disciplines bracketed under “science”, this is not a globally accepted grouping.

Seventh, I believe that, at present, although there is not a focus on a single business model, the implementation guidelines only mentions one: APCs. If the cOAlition is serious about a diversity of business models, then these should receive equal mention, with conditions attached. Further, the cOAlition should consider how its funding streams could be supplied to organisations that run such models. At the moment, there is no accepted way of supporting these organisations when most project funding discretely parcels out outputs and funding into APC-sized chunks. What is the APC-equivalent mechanism for funding non-APC models in Plan S? These models are critical in disciplines where much work is conducted on a non-project-funded basis and also where substantial cost-concentration of APCs can apply. Such models could also, though, provide a solid route to transformation for Learned Societies. See: https://eve.gd/2018/01/21/how-learned-societies-could-flip-to-oa-using-a-consortial-model/ . In the name of transparency, I would note that I have an interest (that may be conflicting) in such models, as I founded and run the Open Library of Humanities.

Eighth, I believe that the implementation guidelines and notes should stress the seriousness of the green, zero-embargo approach as a route. This is a really powerful driver for those Societies or others who feel unable to alter their business models immediately to finally do away with embargoes. I note that several publishers, including Cambridge University Press, already have zero-embargo deposit for scholarly articles in the humanities disciplines but have not seen a collapse in their model.

Additional points, added 2019-01-23:

Ninth, although I understand that monographs and book chapters have been deliberately postponed due to infrastructural, social, and technical challenges, I would welcome clearer guidance on when we might hear on this and what a potential timetable might look like. Books have a much longer lead time than journal articles and it is nearly certain that books are now going under contract that may not be out until 2025. Given this lead time, and given also, as above, that many do not engage unless a concrete deadline is set, it would be helpful to know this. Without such a deadline, there will be no impetus to proceed (despite the fact, also, that setting such a deadline will be accompanied by predictable and inevitable backlash).

Tenth, at present there is an ambiguity around United Kingdom Research and Innovation (UKRI)’s involvement in cOAlition S on which I would appreciate greater clarity. Research England, which manages the UK’s Research Excellence Framework (REF), is part of UKRI. As such, I would expect its future funded outputs (REF submissions) to be subject to Plan S conditions. However, the REF is also jointly owned by the funding organisations of the devolved administrations (HEFCW, SFC, DfE in Northern Ireland). This substantially complicates an understanding of the scale and scope of Plan S within the English funding context and Learned Societies are responding differently, based on their understanding, or otherwise, as to whether REF is involved in Plan S.

Eleventh, given that DOAJ, DOAR, and Sherpa/Romeo are mentioned in the implementation documents, I would appreciate clarity on whether cOAlition S intends to financially support these infrastructural initiatives. Without their ongoing permanence and financial stability/resourcing it is difficult to have faith in these as long-term usable platforms.

Twelfth, the wording of “at the time of publication” in clause 10.1 should be clarified. Having some tolerance here – to allow for, say, illness or researcher unawareness that a piece has actually been published – would be helpful. The three-month window specified by Research England in its green mandate (albeit, from acceptance) could be a helpful clarification.

Thirteenth, I would urge a more distinctive clarification in the document around the word “quality”. When referring to the publication venue, the “quality” in terms of “the quality of academic material” should not apply at the venue/journal level, as DORA and other declarations have stressed. I understood “quality”, here, to refer therefore to the soundness of the infrastructure, digital preservation etc. It would be helpful to clarify this phrase throughout.

Fourteenth, it would be helpful to know whether existing off-the-shelf repository solutions – eprints, PURE, etc. – are compliant with the requirements set out for repositories. Further, features such as “automatic manuscript ingest” should be clarified. How “automatic” should this be, for instance? Is manual oversight by a repository manager acceptable, or is the intention here that this be fully automated?

Fifteenth, for several reasons “Direct deposition of publications by the publisher into Plan S compliant author designated or centralised Open Access repositories” has been resisted for many years by funders. One of the core reasons here is that changes in researcher practices, behaviours, valorization, and understandings of open access have been deemed crucial. Another is that there is a justified fear around deposition infrastructure being controlled by third-party entities with differing sets of motivations and attitudes towards open access. As I understand it, this is why green policies such as the Research England REF mandate avoided such technological solutions (while also being supportive of open initiatives such as the Jisc Publication Router). I would just like to draw this potential area to your attention.

Sixteenth, I believe that greater clarity is required around the right to pursue defamation suits for wrongful attribution under the Creative Commons licenses specified in Plan S. Indeed, I would recommend that this clause be strengthened: “CC BY 4.0 demands that licensees indicate if changes are made when re-using licensed material, and this means that the CC BY-ND license should not be necessary for due protection of the rights of the author. For the protection of authors’ legal and moral rights to published material cOAlition S refers either to the respective Rules of Good Research Practices or to the Berne Convention for the Protection of Literary and Artistic Works.”

There are continuing arguments around misuse of work under CC BY licenses, particularly from the HSS disciplines. This mostly centres around the fact that although the license demands that it be clear that the work has been modified, it does not demand that re-users signal how it has been altered. The concerns here from historians, in particular, centre around political re-use of their material outside the academy by extreme political groups. They are concerned that their words will be altered and attributed to them, by, for example, neo-Nazi groups, with only a footnote specifying that the work has been changed, resulting in reputational damage and historical distortion. Prominent libel suits, such as David Irving v Penguin Books and Deborah Lipstadt over Holocaust denial, indicate that there are consequential and important uses for public history that can result in problems that require recourse to legal remedy.

I believe that some of these fears could be laid to rest by commissioning and publication of solid legal advice on the extent to which defamation and libel suits remain viable with respect to work under the attribution clause of the CC BY 4.0 license. The license, for instance, requires that creators waive their moral rights in order for the rights granted by the license to be exercised. The license allows modification and requires attribution. It therefore makes sense that a modification must be attributed unless waived, albeit with modification noted (but there is no requirement to notify the author of modified attribution). It is possible, then, that reputational damage/defamation could ensue from such attribution but that an author would have waived the moral right to pursue such a claim (the right “to object to any distortion, modification of, or other derogatory action in relation to the said work, which would be prejudicial to the author’s honor or reputation”). It is also possible, though, and this is my belief (although I am not a lawyer and could be wrong), that defamation rests separately from these matters of moral rights in many jurisdictions. The question might hinge, though, in a court case on whether the attribution was wrongful if it indicated that the text had been modified (“I hate Plan S” - Martin Eve, wording modified from original). If this were the case, though, the CC BY license might allow the attribution of anything to anyone, which seems unlikely to be held up in court.

Although I have previously been a major supporter of open licensing for scholarship and research – and still believe it has many, many beneficial effects (and openly license all of my own work) – clarifying this matter seems important. It could resolve many objections in HSS communities where it is felt that this is of real-world import.