SUMMER SERIES: HOW SCIENCE WORKS In the 1990s, the internet upended the scientific publication process. Today, it’s the whole culture of research that’s changing.
The COVID-19 pandemic sent shock waves around the world in the spring of 2020. It also unveiled deep underlying societal and cultural trends. The scientific community rose up and mobilized in the first few months, sharing new discoveries in a fervent attempt to unlock the secrets of the new coronavirus. Everyone was driven by a desire to find a solution - quickly - working under a common research agenda. Clinical data, along with advancements in genetic sequencing and imaging, were disseminated freely among research centers and hospitals, outside the conventional research-publication channels. "It was a case study that illustrates how science actually works today," says Gilles Dubochet, the head of EPFL’s open-science unit. "The model of the scientist who conducts research alone in a lab, coming out only once the results are ready to be shared with a small circle of peers, is steadily becoming outdated. Today, research programs are collective efforts that involve an entire community and produce findings that can be viewed by anyone."
That’s, in essence, the open science revolution: the drive to update research practices so that they’re better aligned with the needs of citizens and the scientific community, and to take full advantage of the possibilities being offered by new technology. The conventional system for publishing research was the first to be upended by this more modern approach. "The current publication process is simply an extension of the model that arose in the 17th century," says Dubochet. "Back then, there were clubs of scientists who discussed things with each other and shared their discoveries through letters. These letters were then published in trade journals so that the discoveries could be communicated to the scientific community." As the years passed, specialist publishing houses took over the role of these clubs, enabling scientists to disseminate their findings through both generalist and field-specific journals. The higher the journal’s impact factor, the greater its prestige. The publishing houses also formalized a system whereby articles submitted for publication were reviewed and vetted by peers. At the same time, university libraries began paying for subscriptions to these journals so that their schools’ researchers could have access to them. As the number of these journals steadily increased, so did the amount of money needed to maintain the subscriptions.
We’re working hard to set up the right mechanisms, and there’s still a lot to do. But the transition is under way.
The arrival of the internet in the mid-1990s changed all that. "Suddenly, anyone could share information pretty much for free, at any time, on any subject, and with everyone," says Dubochet. Scientists, whose research is often publicly funded, seized this opportunity to distribute as much data as possible - from initial hypotheses to end results - to as broad an audience as possible, from laymen to leading experts. That’s what leads the open access movement.
A transition is under way
The challenge in implementing open access, and the open-access policies that make them possible, is finding a way to transform publishing houses’ business models so that they can provide this kind of free, widespread distribution. Universities are in the process of negotiating the details of such a transition by shifting the cost of access to the publication side. "We’re working hard to set up the right mechanisms, and there’s still a lot to do," says Isabelle Eula, head of the EPFL Library. "But the transition is under way."
Public-sector funding agencies are increasingly requiring scientists to publish their research in open-access format in order to be eligible for funding. That marks a big move away from the method typically used to evaluate scientists and research labs, which is based largely on the number of citations they have and on the impact factor of the journals they’re published in. Journals with the highest impact factors (i.e., the most prestigious ones) sometimes don’t offer open access, or offer it only at high prices. "Open-access publications are a way for scientists to boost their visibility," says Béatrice Marselli, the coordinator of the EPFL Library’s publication support team. "They’re also of great benefit to countries and universities that can’t afford subscriptions to scientific journals."
Scientific discoveries are a public good whose value grows as they are adopted by more and more people
In addition to the advantages for individual scientists, open-access policies help advance research as a whole. That’s because the idea is for scientists to share not just their main findings, but the whole process that got them there, from the details of their method to all the data they collected. "The methods that scientists use are still often a black box," says Dubochet. "Even though research papers usually contain a chapter on method, most of the time, the descriptions aren’t detailed enough to replicate a scientist’s experiments. You need the code they used or other information." Unpacking the black box so that other scientists can perform the same experiments is called making the research reproducible. "We need to set up an entire research system where the emphasis is on open access and effective collaboration," says Dubochet.
Opening up in both directions
Open-access publications are just one element of a broader trend towards open science - that is, the sharing of scientific tools and knowledge through an array of new channels that run alongside standard research publications. These channels include online platforms for sharing data, code, partial or not-yet-peer-reviewed results, and papers presenting research failures where the results didn’t come out as expected. This latter practice has actually been used by physicists for years. "Publishing the results of research that doesn’t turn out as hypothesized is just as important as publishing work that proves your hypothesis right," says Marselli. "Both types of knowledge are essential for moving science forward." Open science therefore calls for a rethink of the existing peer review process and the criteria currently used to assess research results.
In the long run, isn’t open science really just about the duty that the scientific community has towards society? "Scientific discoveries are a public good whose value grows as they are adopted by more and more people," says Dubochet. "With open-access publications, doctors and journalists can read a research paper first-hand to inform their decisions about how to treat a patient or the best way to report on a new advancement to the general public. Open science takes this one step further, disseminating not just an end conclusion but each step of the path along the way. All this helps the scientific community strengthen - or rebuild - its relationship with society, because the opening up has to go both ways."