Forums

EDITORIAL

JAKE S. YESTON , VALDA VINSON , AND H. HOLDEN THORP Authors Info & Affiliations
SCIENCE
5 Jan 2023
Vol 379, Issue 6627
p. 7
DOI: 10.1126/science.adg4111

The new year 2023 arrives with promising developments in fusion research. In December, scientists at the US National Ignition Facility (NIF) focused 2.05 megajoules of laser light onto a capsule of fusion fuel and produced 3.15 megajoules of energy. This was the first laser-driven fusion demonstration in which the reaction produced more energy than the laser light used to start it—a goal set for NIF at its founding. Massachusetts Institute of Technology physicist Anne White, who was not involved in the work, described it to Science as “a major breakthrough.” After a year of ups and downs in science policy and research, this was the kind of exciting event needed to propel scientists into the promises of 2023.
As always, with these kinds of advances, scientists and the federal government must be careful not to exaggerate the practical implications. The achievement was announced at a high-profile press conference hosted by the US Secretary of Energy. But as Science’s Daniel Clery wrote, “fusion power stations are still a distant dream.” The NIF result is a glimmer of promise, but many technical challenges still lie ahead. Scientific results announced by press conference prior to the publication of a peer-reviewed paper do not always hold up well, so we will have to wait to see the data, hopefully this year.
Similarly, 2023 brings new energy to Science. We are making a number of important changes to our Information for Authors and Editorial Policies. These include updating the types of research papers that we publish and adding a new category for retractions.

Over the history of the magazine, research papers have gotten longer and longer with more data and figures. These developments are good for open science, as the experimental methods and results must be reported in sufficient detail to allow reproducibility. As a result, Science is ending the category of Reports, which were shorter than Research Articles. Now there will be one category of research paper—the Research Article—which will be around five pages in the printed journal. This will typically include 2000 to 3000 words of text, three to five figures, and 50 references. The majority of Science’s papers already fit into this category. This format still requires the focused writing that enhances the clarity of the published papers, but without compromising transparent reporting on the work. Supplementary Materials will continue to be published online, with the accompanying data, curated at repositories such as Zenodo and Dryad, expected to grow. We will continue to allow some longer Research Articles to appear online and with a one-page summary in the print magazine.
Up until now, two criteria for retractions have been explained in the Editorial Policies. One applies to instances when misconduct is clearly demonstrated, and the other concerns errors that undermine the core conclusions of a paper. All other errors were often accommodated by Science’s publication of an Erratum or a Corrections notice. We are now adding a third criterion for when a paper has received enough corrections or contains enough errors to cause the editors to lose confidence in it. Although Sciencehas always had the latitude to retract a paper for reasons deemed sufficient to the editors, we are now stating this additional criterion explicitly.
The analysis of published images in research papers has become more prominent and cataloged on sites like Pub-Peer, and image sleuths like Elisabeth Bik have brought more energy to the careful maintenance of the scientific record. Scientists are human, and mistakes—whether intentional or inadvertent—are inevitable by both authors and journals. Fortunately, science is a self-correcting process that ultimately leads to the right answer. Unfortunately, when we contact authors about concerns that have been raised about their Science papers, we are often met with defensiveness and denial. That needs to change. A quick correction or retraction builds confidence that we all are committed to an accurate scientific record. The community needs to work to reduce the stigma of correcting the record so that—in this time of extraordinary scrutiny and waning confidence—we can show the world that self-correction provides a basis for trust. We hope that this new criterion for retractions will contribute to enhancing that trust.
Here’s to an energetic 2023—for fusion and more.