Eighteen ways of looking at metascience
Eighteen ways of looking at metascience
Today in Barcelona we launch Seize the Meta-Moment: the most comprehensive account yet of how the science of science is organised, funded and governed around the world. Here's what we found.
By James Wilsdon, Josie Coburn and Seunghyun Lee (RoRI)
SubscribedThis morning, under the concrete vaults of CosmoCaixa, Barcelona’s museum of science, we pressed the button on the biggest project RoRI has ever undertaken. Seize the Meta-Moment is out: a report running to more than 250 pages, eighteen country case studies, and an open, interactive atlas of nearly half a million publications. Colleagues who wrote the case studies have travelled from Bogotá, Seoul, Nairobi, Wuhan, Canberra and a dozen other places to present their own systems, in their own words, to a room full of the funders and researchers who make up the RoRI consortium. The head of the US National Science Foundation’s weeks-old Office of Metascience joins us by video this afternoon to close the day.
That cast list is the first thing to understand about this project. For a year, teams in eighteen countries — Australia, Austria, Brazil, Canada, China, Colombia, Germany, India, Ireland, Japan, Kenya, Korea, Namibia, the Netherlands, Norway, South Africa, the UK and the US — have been mapping how their research systems study, evaluate and reform themselves. Each team wrote its own account, against a common framework but in its own vocabulary. We then set those accounts alongside two independent data-driven maps of the research literature: one built from the labels that metascience communities use for themselves, and one grown outward from more than a thousand publications the country teams told us mattered. Nobody has attempted a comparison of national metascience systems on this scale before.
Why now?
Because metascience is having a moment, and the moment is accelerating. In the year since this project began: the UK trebled the budget of its Metascience Unit, to £49 million through 2030. The White House published Science: A New Golden Age, calling for “an empowered metascience unit” in every federal science agency, and NSF launched its Office of Metascience in September. In October, ministers and senior officials from seventeen countries endorsed the Kyoto Vision for a Golden Age of Science — the first multilateral ministerial statement to welcome metascience by name. Japan wrote metascience into its 7th Basic Plan. And on 3 December, Research Ireland launches its new meta-research unit, with the UK and US units (and RoRI) in the room.
Attention of this intensity is a gift and a hazard at once. It brings money, mandates and legitimacy. It also brings the risk that a rich, plural, decades-old set of traditions gets flattened into a single story — usually an Anglo-American one, about efficiency, bureaucracy and “fixing” science. Our report is, among other things, an argument against that flattening.
What we found
Three findings run through everything else.
First, the label is new; the activities are not. The world’s first journal devoted to the study of science appeared in Poland in 1918. China’s science of science community has been at work since 1982. Austria’s evaluation platform is thirty years old; Brazil’s tradition of studying science and technology goes back seven decades. In Korea, 23 of our 31 survey respondents reached first for “STI policy”, not metascience. What is genuinely new is the travelling label, the international networks and the appetite for experiment. A report that treated the meta-moment as a new arrival would misdescribe more than half its cases.
Second, metascience is far bigger than its name. Searching for the eight communities that identify with the field retrieves around 59,000 publications. Our atlas — grown from the papers the country teams themselves nominated — contains 476,154 records across 131 topics, and 86 of those topics include none of the papers the teams supplied. The field is more than ten times larger than any label captures, and much of the work is done by people who would never call themselves metascientists. That has practical consequences: if you only fund and convene the people who wear the badge, you miss most of the capability.
Third, there is no single right way to organise it. We find five recognisably different national models: metascience directed by the state (China, Japan, Korea, India, Colombia); delegated to funders (Austria, Canada, Ireland, Norway, Namibia); given a dedicated institution (so far, uniquely, the UK); left to decentralised communities (Germany, the Netherlands, Brazil, Australia, Kenya, South Africa); or driven by entrepreneurial and philanthropic actors (the US). These are configurations, not grades. Countries combine them and move between them, and each has characteristic strengths and a characteristic gap. The report’s final chapter suggests where each might usefully begin.
Around those headlines sit findings that deserve more debate than they usually get. Most of the money in metascience is invisible: twelve of our eighteen case studies contain no investment figures at all, and where figures exist they are small — the UK’s £49 million is roughly 0.06% of its public R&D spending. Many systems cannot see themselves: seven of the eighteen lack an integrated national research information system, and those with superb data do not necessarily learn from it (”strong data infrastructure alone does not automatically lead to mature metascience capabilities”, as our Korean colleagues put it). Capability rests on too few people, sustained by volunteers, secondments and the occasional solitary crusader. And metascience is political — its vocabulary has been adopted at the highest level of the US government in ways its originators “may not be able to control”. The answer is not to pretend otherwise, but to make the politics explicit: a condition of learning, not a defect.
Why this matters for RoRI
RoRI was founded in 2019 on a simple idea: that transforming research systems needs robust, comparative evidence, and that funders — who hold the data and control many of the levers — must be partners in producing it, not just consumers. Seize the Meta-Moment is that idea made real at full scale. It was commissioned by our partnership of funders, co-chaired from Vienna and London, written with teams on six continents, and built so that any funder, ministry or research community can locate their own system within it: which capacities they already have, which they lack, and who else has faced the same problem.
It is also, deliberately, a resource rather than a verdict. Everything is open: the report, the case studies, the search queries, the atlas. Where our bibliometric methods capture a community poorly — a caution our South African and Chinese colleagues rightly press — we say so, and publish the material so others can do better. Over the coming weeks, we’ll hand this space over to the country teams, with a series of Go Meta! posts digging into individual systems. Consider this post the overture.
Making the moment count
The report ends with recommendations: for the governments, funders and organisations using metascience, and for the academic communities contributing to it. Underneath them all sits one foundation — co-creation — and one warning. Moments pass. Political attention, new money and a shared vocabulary have arrived together, and they can dissipate together. Seizing this moment cannot mean building the largest possible apparatus while attention lasts. It means investing in what remains useful when attention has gone: data that are linked and open, routes from evidence to decisions, careers and institutional memory, and the habit of testing alternatives and publishing the results, flattering or not.
The final chapter of our report opens with Brutus, on the eve of battle: “On such a full sea are we now afloat, and we must take the current when it serves, or lose our ventures.” The current is running. Eighteen countries’ worth of evidence on how to take it is now yours to use.
Seize the Meta-Moment: How metascience is evolving worldwide is available open access, with all eighteen country case studies and the interactive data atlas, at meta-moment.researchonresearch.org. The accompanying working paper, Research on Research in Africa: Mapping the Metascience Ecosystem, by Similo Ngwenya, is also out today.
James Wilsdon is Executive Director of RoRI; Josie Coburn and Seunghyun Lee are Research Fellows in Metascience at RoRI. Together they coordinated the Meta-Moment project.
Thanks for reading Go Meta! Subscribe for free to get essays, reflections and provocations on how research is funded, evaluated and governed.
© 2026 Research on Research Institute
RoRI CIC, C/O STEaPP-UCL, 4th Floor, Shropshire House
11-20 Capper Street, London WC1E 6JA
UK