UNESCO Recommendation on Open Science – Thoughts by Barend Mons

Prof. dr. Barend Mons
LUMC
Scientific Director GO FAIR Foundation
President of CODATA

I would like to invoke a classical piece of wisdom to focus my reflection on what is commonly referred to with the all-embraceable hype term “open science”: namely, “putting new wine into old wineskins.”[1] Quite often, the concept of open science is immediately reduced to the idea of open access articles, which is only a minor aspect. Fortunately, the UNESCO recommendation includes serious paragraphs about machine-actionable FAIR (meta)data, workflows, and other research outputs in addition to traditional scientific narrative. Still, terminology like “multilingual scientific knowledge” seems to support the outdated concept that knowledge and evidence-based information should be primarily communicated via human-readable narrative. However, we must realise that the evidence on which we base our knowledge should be centered on data and relevant, reproducible, observations and patterns that lead to precise claims[2], rather than on storytelling. Narrative is necessary but is supplementary to data and actual claims.

 We have entered a disruptive decade, in which data is growing so exponentially that incremental changes in the current system of scholarly communication are doomed to fail. The original meaning of “putting new wine into old wineskins” is “do not put new (still-fermenting) wine into old, (dried-out and resistant) wineskins, as they will burst and spill the good wine”. I would like to classify not only the UNESCO recommendations but also the recommendations it builds on[3] as delicious-tasting “new wine”. What is especially tasty about the UNESCO recommendations (and those of the G20, for example) is that (FAIR) (meta)data and infrastructure are clearly mentioned, and open access to articles is not presented as a proxy. There is also an emphasis on training, proper data stewardship, and even explicit machine readability and actionability. Although I still have harsh questions regarding implementation after reading the document, the fact that there is now a comprehensive, UN-level document to refer to is of high symbolic value. Many policy makers and universities will likely start their own fashionable, open science declarations with a reference to the UNESCO recommendations. But, so far, most continue to put this still-fermenting new wine into the old wineskins of their current reward systems and publishing requirements. Ultimately, the escape from the 17th-century scholarly communication prison is not about blaming the publishers, but about facing our own, dried-out, elitist, and anachronistic ivory-tower scholarly communication practice (from which the publishers live lavishly).

 All the beautiful, sustainable-development-goals rhetoric and “leaving no-one behind” language trigger cynicism in my now old and repeatedly reset and battered brain. Of course, we wholeheartedly agree with all these velvet notions, but do we not all continue to fill old wineskins, by failing to reward our postdocs for sharing data, by not sharing vaccines, by not sharing our knowledge or our wealth, and by robbing less fortunate people of their playful youth and natural resources so that we in the Global North can have our electric cars and cleaner cities? Why would science be different? The (almost) universally agreed-upon (among intellectuals) new wine, although wonderful and tasty, goes quickly into the old wineskins of the current, journal-based scholarly communication and reward system, which will resist until it finally bursts. Many universities, while readily signing all declarations on open access and beyond, meanwhile only appoint professors based on metrics based on a Journal Impact Factor element, so they keep themselves in the deadly embrace of a publishing system that never really left the printing-press age, including the exorbitant prices that simply switch from readers to authors, to the detriment of underprivileged researchers. 

 Even more fundamental: In the time of COVID-19, if science showed us anything beyond reasonable doubt, it is that narrative articles lacking machine-actionable supporting data that can be re-analysed and scrutinised only benefit the few researchers who have both access to it and the luxury of a workforce to find the (few) gems in this debris-filled literature avalanche. We all know the figures: hundreds of thousands of articles on COVID-19 (the top day for our text mining colleague counted 450 articles) and currently over 4 million hits in Google Scholar. In many cases, the underlying data, while understandable by the elite in top institutions, are hardly accessible by the general public. Even if it is mostly open access, text remains the proverbial nightmare for machines. Constructing a proper, machine-actionable model of COVID-19 disease progression based on this crippled scholarly communication system was a sobering experience, I can tell you.

 To end on a positive note, in the funding and publishing worlds, there are strong movements[4] towards requiring and supporting publication practices that will make computers our strong allies in science. The open science cloud/commons initiatives around the world are starting to collaborate and converge, and there is an ongoing global effort to define the minimal standards for an “internet of FAIR data and services”[5]. Revolutions take time,[6] but we do not have much time to waste on certain sustainable development goals.   “Entrusting UNESCO with the mission to coordinate, in consultation with Member States and relevant stakeholders, the development and adoption of a set of open science goals” is not possible without executive mandate and power. My proposal would be for UNESCO to use its established collaborations with the International Science Council and its data organisations (CODATA and WDS) in the context of the DataTogether collaboration,[7] to ensure that the declarations in UNESCO’s recommendations move from symbolism to execution.

[1] This phrase was originally coined in the New Testament in this subject-predicate-object sequence, and is proof of rapid miscommunication and conceptual drift in human language. Many use the saying “old wine in new wineskins”, which has the opposite meaning and is used to mean dressing up “old stuff” as new.

[2] This includes precise formulation of these claims in non-ambiguous, preferably FAIR, formats.

[3] (EOSC, G20, G7, OECD, ISC, FAIR principles, etc.)

[4] DOI: 10.3233/ISU-200102; https://content.iospress.com/articles/information-services-and-use/isu200102

[5] www.fairdo.org

[6] https://www.mdpi.com/2078-2489/12/11/472, Information 2021, 12(11), 472; https://doi.org/10.3390/info12110472

[7] https://codata.org/wp-content/uploads/2020/03/Data-Together-Final-Version-March-2020-FINAL.pdf