Who discovered the Higgs boson?
It’s been three years since physicists at the Large Hadron Collider, the atom smasher near Geneva, discovered the Higgs — or, as the press calls it, “the God particle.” Since then, scientists at the collider have emitted a torrent of publications, each including hundreds or thousands of authors, listed alphabetically. Many of the papers have been led by one name: “G. Aad.”
The God particle. G. Aad. Could it be a coincidence, some scientists have wondered, or was this a joke, or a way of diminishing academic jealousy as hundreds of early-career physicists enter the job market? Each could claim credit for the discovery. But just how much?
The idea of scientific authorship has been under strain in recent years, and not just from physicists. For nearly all disciplines today, team science is the rule. These teams have grown increasingly large, including consortia that lead to hundreds of authors being named on a single study. Yet scientific papers, when assigning credit, still use a simple formula: a list. For the primary authors on these papers, who garner renown for a study, such a system can work out just fine. But other scientists, many possessing vital talents, have found themselves lost in a sea of co-authors, their contributions obscured and mysterious to tenure committees.
It’s a growing problem, one that defies simple fixes. “The people in the middle are essential, and they don’t get recognition,” said Arturo Casadevall, a professor at the Albert Einstein College of Medicine at Yeshiva University.
Biomedicine, with its glut of early-career researchers and recent genomics-driven emphasis on teams, is ground zero for this dilemma. Many researchers specialize in bioinformatics or biostatistics, for example, and their sophisticated analysis can be essential to every paper published by a team. Yet, unless they are canny in arranging how credit is apportioned, they could easily see themselves without a lead authorship when the team’s suite of papers is released.
“It can be very intimidating, and, in some sense, you can tend to get lost,” said Anshul Kundaje, of life inside the consortium.
Mr. Kundaje would know. Now an assistant professor of genetics and computer science at Stanford University, he spent years as a postdoc and staff scientist contributing to the Encyclopedia of DNA Elements, a large biology consortium. By focusing on his particular hypothesis within the consortium, Mr. Kundaje was able to carve out independence enough to be the lead author on a paper, impressing potential hiring committees. But many of his peers have had trouble securing jobs, and receive little credit for the important-but-invisible contributions they’ve made, he said.
Many computational biologists, especially, end up in insecure grant-supported positions and eventually leave academe for business, added Daniel S. Katz, a senior fellow in the Computation Institute at the University of Chicago. Or, if they follow a common line of career advice, they pursue their own work and avoid teams, even if it means duplicating efforts by others. There’s more redundancy, but everyone gets a lead authorship.
“This is what kills me,” said Cassidy R. Sugimoto, an assistant professor of library and information science at Indiana University at Bloomington. The crediting system is “based on previous models of science, and science isn’t like that anymore.” If these researchers are “truly critical to science,” she continued, “then we should value them. And I don’t think we’re doing that in the current system.”
Teams aren’t going away. The Large Hadron Collider, though an extreme example — one recent paper from it, in May, topped 5,000 authors — better represents where science stands today than the romantic vision of a lone scientist in her lab. Teams now conduct more than 90 percent of scientific research, and it’s common to see papers with 10 or more authors. Financing agencies push for large, interdisciplinary projects, and the complexity and computerization of science has forced researchers to possess ever smaller slivers of its expanding edge.
New Approaches
While reforms have been slow, experiments are under way. Some universities have adjusted their tenure criteria. Journals have tweaked how they describe author contributions. And, most promisingly, one project is seeking to revolutionize how scientific contributions are recognized, potentially doing away with the simple author list once and for all.
It’s work that’s overdue, Ms. Sugimoto said. Even departments like history or philosophy are grappling with how to divide credit among teams of researchers brought together by digitization.
“Every discipline has to face this,” she said. “Even in the humanities.”
When Ronald N. Germain, an immunology researcher at the National Institute of Allergy and Infectious Diseases, asked a decade ago to establish a department of systems biology — a discipline rooted in networks — he envisioned not a group of principal investigators pursuing their own goals, but rather a team applying complementary skills.
This presented an immediate problem. According to how the National Institutes of Health judged tenure, his team members, each untying one end of the scientific knot, would have trouble securing their careers. “I found it unethical to recruit them if they couldn’t advance,” Dr. Germain said. Rather than giving up, though, he fought to change the system.
Now when NIH researchers involved in team science come up for review, they are given the same credit for a middle authorship as for a senior authorship, Dr. Germain said. “You don’t sacrifice and don’t fight endlessly,” he said. Dr. Germain has had several team members come up for review since the change was made. It worked: Creation of software was an explicit part of one promotion, he said.
The NIH changes are often cited as a model for universities, but Dr. Germain still finds many researchers unaware of them. He’ll appear at a meeting, describe the review process, and end up showered with business cards. “You’ve really managed to get that change in writing?” people say. “And it really works?”
Some universities have improved how they evaluate team science, though exact figures are fleeting, said Brian Uzzi, a professor of leadership at Northwestern University and co-author of a report on team science published this year by the National Research Council. But the fixation on lead authorship as signifier of intellectual independence remains. This is understandable, he said. What other metric could they use?
One new source of insight could come from better defining the roles of each researcher in a paper. That’s the idea behind Project Credit, an effort led by Liz Allen, head of evaluation at the Wellcome Trust, the influential British biomedicine foundation, and Amy Brand, incoming director of the MIT Press.
Several years ago, they saw a growing divide between the glowing recommendations written for some scientists and their persistent lack of lead authorship. This indicator was clearly failing. Was there another way?
Inspired by a few journals, like those published by the Public Library of Science, that had begun classifying authors by their contributions in basic categories, Project Credit has developed a taxonomy of 14 roles that scientists may play in a team research project. Each contributor can hold multiple roles; over time, as the structured data follow a researcher, these categories will allow an individual’s talents to stand out. Listed in full, they are a précis of the scientific enterprise: conceptualization; methodology; software; validation; formal analysis; investigation; resources; data curation; writing, original draft; writing, review and editing; visualization; supervision; project administration; funding acquisition.
Right now, Project Credit is still a pilot. The team is in talks with a producer of manuscript-writing software to include its taxonomies, and Cell Press, one of the scientific world’s most influential publishers, has begun urging researchers to use the categories on submissions. Mozilla Science Lab, an open-source science association, has also used the categories as the basis for its contributorship badges, a way of visualizing the accrual of scientific skill.
‘A Tipping Point’
The field of physics has grappled with team science for decades — ever since the Manhattan Project.
“There’s a tipping point where the papers can’t reflect the contributions the authors made,” said Michael S. Lubell, a physics professor at the City College of New York. “In the case of particle physics, it’s way beyond the tipping point now.”
Even so, the vast team behind the Large Hadron Collider has proved to be a challenge. In 2008, the International Union of Pure and Applied Physics warned that it was becoming difficult to assess individual achievement, and recommended seven steps to make contributions more transparent. Those recommendations have not been followed, said Gregor Herten, a physicist at the University of Freiburg who led the report.
“I do not blame anybody, because this is a difficult problem,” Mr. Herten said. Teams require communication and trust. It’s crucial that improved transparency does not lead to the kind of fights over authorship seen in biology, he said. “It would be very destructive to have endless discussions for each paper on who are the major authors.”
For now, physics will muddle on, highlighting early-career researchers by talks at conferences or word of mouth, much as it always has.
As for G. Aad? Georges Aad does exist, thank you very much, he wrote in an email. A scientist at a French particle accelerator, Mr. Aad is an expert at identifying b-quarks out of the zoo of exotic particles tossed out of the collider. It’s been a perk topping the collider’s papers, he said — people do remember his name. Friends also tease him as the Higgs discoverer.
He also gets the rare letter from theorists outside of the project who are unaware of its alphabetization scheme. You and your team, Georges Aad, they write. You’ve got it all wrong.
Paul Voosen is a senior reporter covering the sciences. Write him at paul.voosen@chronicle.com; follow him on Twitter @voooos; or see past work at voosen.me.