Image Credit: h_pampel,

In the quest to eliminate barriers to research access, scholars are confronting a host of questions with regard to how the academic community can affordably and sustainably make content freely accessible to readers. To this end, many are also considering if and how the way research is communicated should change in the digital age.

Ulrich Herb is no stranger to such inquiries. One of the first German signatories of the Budapest Open Access (OA) Initiative, Herb has been a part of many OA projects including working as project manager and open access expert at Saarland University and State Library and serving on the Open Science Working Group of the Knowledge Foundation. He is also completing a forthcoming book with co-editor Joachim Schöpfel, Open Divide? Critical Studies on Open Access, which will bring together seventeen critical studies of scientists and librarians involved in OA projects throughout the world to explore expectations and realities of OA in the fifteen years following the Budapest OA decision. In a recent interview with Scholastica, Herb shared his thoughts on what he foresees as requirements for a sustainable OA future, including key areas where he believes journal publishing must be improved.

Q&A with Ulrich Herb

You are part of multiple OA and electronic publishing working groups. What are the main challenges and opportunities for open access and digital publishing that these groups have identified?

UH: This question first relates to electronic publishing in general, which is - apart from some disciplines from the humanities that stick to printed publications - state of the art. Unfortunately, in most cases electronic publishing means not much more than producing and using PDFs that do not really offer many added values compared to printed articles or books. The next step is already being taken in some areas: Enhance electronic documents with features that are unknown to printed publications to create interactive or living documents. Such documents could be produced in a collaborative manner using git or subversion systems known in software development. These systems also offer the opportunity to track micro contributions to text publications and allow credit to be given to persons who usually never appear as contributors in printed publications.

Nevertheless, one should not forget that digital objects may be subject to restrictions that are unknown to printed publications. For example, while I may read a printed text as often as I like, lend it to someone, or do whatever I want to do, digital texts may be subject to usage restrictions that interdict all these actions. In a nutshell: The value of digital objects is mainly a function of the conditions of use that apply to them. The more open these conditions are the more value an object may offer - of course depending on its content. Scientific results show that the share of OA publications among the body of scientific journal literature is growing constantly and it seems that OA will perspectively outnumber closed access. The main reason for the success of OA journals might be that scientists are now aware of its benefits, such as higher citation counts; funder mandates are requiring scientists to publish OA; and commercial publishers consider OA no longer a threat but a lucrative business model.

Do you think the APC model is scalable? What pros and cons do you see in journal funding shifting from subscriptions to APCs?

UH: It depends. If Gold OA is dominated by commercial publishers the APC model will probably generate costs above the costs known from the subscription model. The rules will be the same as in the subscription world where in the year 2013 Elsevier, Springer, and Wiley-Blackwell produced more than 47% of the scientific journal literature. This concentration gives them the power to dictate prices. Apart from that, funding agencies are paying APCs that are far above the cost of producing an article. So every commercial publisher that does not raise its APCs would be a little stupid.

In 2015 Emerald, for instance, raised the APCs of 32 engineering and technology journals from $1,595 to $2,695, or nearly 70% per paper. When Richard Poynder asked the publisher why they increased the prices, the answer was more or less, because compared to journals with a similar scope and reputation Emerald’s APCs were too low. Furthermore, journals are not substitutable as they differ regarding reputation, Impact Factor, and scope. There won’t be a true market so long as scientists don’t have freedom in choosing where to publish. Currently, they have to publish in journals with the highest reputation and in the long run these journals will be the most expensive ones - supposing they are in the hands of commercial publishers. I guess APC-based Gold OA will only be sustainable if the journals are offered by non-commercial publishers.

From following the OA movement, do you have an idea of what rights not-for-profit presses give up when they parter with corporate publishers?

UH: As far as I know, in most cases the rights concerning the title of a journal are transferred. In fact, I’ve rarely heard of a journal that did not give these rights to a commercial publisher. This is apparent in many cases, such as the editors of Lingua having to break away from Elsevier due to its business model and OA policy and launch a new journal, Glossa. It was the same with the Journal of Machine Learning that was published by Kluwer Publishing (nowadays by Springer Nature) and that had to be newly launched as the Journal of Machine Learning Research. The transfer of rights is surely a relic from the print era when the editors needed publishers to distribute their content as printed issues and had no other choice except to transfer the rights. Nowadays this is anachronistic and gives the chance of blackmailing.

Your forthcoming book Open Divide? brings together critical studies of OA from scholars and librarians around the world acknowledging both pros and cons in OA models - what are some of your main findings? Do you think some issues in the current publishing system are/could be recreated in OA publishing if the system isn’t substantially altered?

UH: As the book is still in the making I can only speak for me. In fact, I think that many of the structural elements known from the subscription market will re-emerge in OA if the predominant publishing system doesn’t change. No matter if we think about open or closed access the scientific publishing field still consists of the same stakeholders with the same power (or without it) and interdependencies. Apart from pricing problems described above we see (and will see) concentration, Springer bought BioMed Central and Frontiers, two OA publishers, and Elsevier is the largest OA publisher in the world.

What do you think is the most important way that journal publishing needs to be improved? Do you think there is an ideal publishing model, and if so what would that look like?

UH: Simply spoken, it would adopt Open Science principles in a sustainable way. Firstly: Cost effective with APCs based on the true production cost of an article. This would also mean that the calculation of first copy costs would have to be transparent. Secondly: Data and software belonging to articles should be available OA. Thirdly: Metrics used to assess scientific information should be open. Fourthly: If scientists want to use an open review (as authors or reviewers) they should have the option that a submission is subject to open review or that the review is published. Review and metrics are important since they don’t merely report on impact of science but strongly steer it as scientists often orient their actions towards such evaluative criteria. Unfortunately, altmetrics are often considered open, but this is a fallacy as most altmetrics services are just as proprietary and closed as citation data from the Web of Science or Scopus.