Making Sure Open Science Stays Open: 10 Years of Advocating Open Science Policy

February 7, 2022

Dr. Jean-Claude Burgelman
Editor-in-Chief of the Frontiers Policy Labs

In the spring of 2011, I gave my first speech on open science at an internal European Commission meeting consisting of a rather small group of interested scientists from the Joint Research Centre (JCR). We were confident that science was facing the same deep changes—from the analogue to the digital—that were occurring in the economic sector and society at large. We therefore called it Science 2.0, to indicate that we expected the change to be of a similar order of magnitude to that of Web 2.0.

Commentary

Making Sure Open Science Stays Open: 10 Years of Advocating Open Science Policy
Jean-Claude Burgelman
February 7th, 2022

In the spring of 2011, I gave my first speech on open science[1] at an internal European Commission meeting consisting of a rather small group of interested scientists from the Joint Research Centre (JCR). We were confident that science was facing the same deep changes—from the analogue to the digital—that were occurring in the economic sector and society at large. We therefore called it Science 2.0, to indicate that we expected the change to be of a similar order of magnitude to that of Web 2.0.

Ten years ago, the impact of digital technologies on science was not widely accepted. In those days, most talk of change in science consisted of heated discussions about paywalls in scientific publications and the need for open and free access. “Open” was equated with “better” and “free”, but from experience I knew that “open” did not automatically mean either. This explains why I tried to avoid the term “open” as much as possible in open science policy debates[2].

Ten years and more than 140 presentations later, I renounced my generic advocacy and delivered a keynote called “The Value of Open Science”[3]. By this time much had changed, and open science was a fact. Open science had evolved from a marginal discussion amongst radical and often sceptical science policy geeks[4] to a central topic in science policy making. The EU played a leading role as the first political entity to propose a comprehensive policy addressing the system change that science was undergoing[5]. In 2021, the mainstreaming of the open science conversation culminated in a highly symbolic moment when UNESCO, with 193 countries as members, issued its recommendation on globalizing open science. Frontiers Policy Labs paid tribute to this milestone here.

Today, making open science work has also become a dominant theme in science policy making. Around the world, thousands of specialists are working on making research data Findable, Accessible, Interoperable and Reusable (FAIR)[6]. Inspired by the huge impact of Plan S, the dominant publishing schemes are being challenged. Most significantly, open science clouds, which are the virtual environments for open science, are being created across the world and executives are meeting informally to discuss the policies needed to establish a globally “interoperable” cloud[7].

I believe open science has proven its value and is here to stay. The system change that we barely detected 10 years ago is now in full swing and widely accepted de facto. So, should we relax now? On the contrary. With every paradigm shift, both opportunities and challenges are created. I foresee both obvious and more subtle challenges ahead for policy makers and open science advocates.

One obvious challenge is the possibility of “fake science” due to abuse of openly available data. I would argue that this is the smallest challenge and easiest to solve. If we use the number of retractions as a proxy (2 of every 10,000 published articles[8]) then we should obviously not panic. Although the possibility of scientific fraud or data manipulation is real, most scientists will use the open science paradigm to do more and better science. If we continue to have solid peer review at the core of our science system—and use AI to help meet the large submission volume—the problem will be manageable and easy to detect.

A more serious problem is that much of our scientific knowledge remains behind paywalls—only a small percentage of global publications are openly accessible. Moreover, a significant part of what is openly accessible is under a transformative agreement, which is an expensive method used by incumbent publishers to make the science system pay for publishing in open-access format. It is remarkable that publishing remains an unregulated market in terms of consumer choice and price setting[9].

In terms of the less obvious problems open science faces, private digital platforms like FAANG (Meta (formerly known as Facebook), Amazon, Apple, Netflix, Alphabet); host or control much of the data that science needs—without playing by all the rules. All cloud providers would like to access scientific data to run and test their algorithms because scientific data are of excellent quality. For example, Amazon could be interested in acquiring a significant publisher of life sciences because of its recently announced intent to roll out an online pharmacy. Or why not Google, given its sister Alphabet companies focused on life sciences? Additionally, the traditional science publishers are increasingly becoming science platforms deploying the same strategies of vertical integration and lock in, via the research data they host or produce, with the recent Clarivate-Pro Quest merger being the latest example[10].

Reflecting on what happened during the digitalization of other sectors of society, we can predict more FAANG-isation of research data, as that is where strategic value-making happens, from a FAANG point of view. Ironically, the universe of metascience, which is possible because of open science, might indeed become meta-versed[SD5] [JB6] .

As a scientific policy community, we therefore need to establish a few guidelines to avoid excessive control over scientific output and enable maximum circulation of the results of science:

·       No more paywalls for scientific publications

·       No [SD8] unintended use of research data

·       Research data should be as open as possible and only as restricted as necessary

·       FAIR principles should be applied to the entire research life cycle

·       Machines must be accountable to men

 Good intentions alone will not safeguard open science. We therefore need a global de facto/de jure recognition that science is a common good of society[11] and permitting the global availability and accessibility of:

•       All publicly funded scientific output

•       All research data produced by the global public science system and related meta data

•       All code used for making sense of these data

•       All workflows to describe research steps and data transformations

•       All research software, algorithms, and tools for reuse

I am quite sure that, in the next couple of years, we will no longer have to argue for the validity of open science as the modus operandi for science in this century. I predict that we will, however, need to concentrate our attention on making sure that open science remains open—and Frontiers Policy Labs will continue to contribute to that worthy goal.

[1] https://firstmonday.org/ojs/index.php/fm/article/view/2961 DOI:  https://doi.org/10.5210/fm.v15i7.2961
[2] Why “open” replaced Science 2.0 is described in J.C. Burgelman (2021) Politics and Open Science: How the European Open Science Cloud Became Reality. Data Intelligence, Volume 3, Issue 1 
[3] https://www.linkedin.com/posts/jean-claude-burgelman-528a7566_open-science-adds-value-to-science-and-the-activity-6875143795659309056-kI0f
[4] https://www.economist.com/leaders/2013/10/21/how-science-goes-wrong (paywalled)[5] This is largely documented at https://www.frontiersin.org/articles/10.3389/fdata.2019.00043/full
[6] https://www.go-fair.org/fair-principles/
[7] https://codata.org/initiatives/data-together/oscer/
[8] https://www.science.org/content/article/what-massive-database-retracted-papers-reveals-about-science-publishing-s-death-penalty
[9] See https://researchprofessionalnews.com/rr-news-europe-views-of-europe-2021-1-scholarly-publishing-needs-regulation/
[10] https://sparcopen.org/news/2021/sparc-statement-on-completion-of-clarivate-proquest-merger/
[11] First proposed at the first UN OS Conference, 2019 https://research.un.org/conferences/media and recently echoed in a LERU data statement https://www.leru.org/files/LERU-Data-Statement_12.2021.pdf