Microsoft Adds Momentum to “Open Science”
Shutterstock / Sergey Nivens
Momentum continues to build behind the “open science movement,” propelling the debate over publication of scholarly works and the scientific process itself.
Last week, Microsoft Research announced it was adopting a policy that allows it to retain a license for research submitted to conferences or publishers in order to post it to a freely accessible online site as well. And earlier this week, pharmaceutical giant Johnson & Johnson said it will release data from clinical trials, through an agreement with the Yale University Open Data Access Project.
The decisions represent at least a subtle shift in the standard corporate impulse to retain a death grip on any research that could potentially make its way into products. To be sure, Microsoft isn’t about to start giving away the raw code behind Office. But they’re loosening the reins on the research division, allowing their published work to reach beyond those who can afford expensive industry journals and conferences.
In J&J’s case, researchers will have to submit requests to Yale for access to the data, but the information will be provided free of charge.
The basic concept behind open science is that knowledge should be as widely disseminated as possible, inviting participation beyond esteemed institutions, private enterprises and tenured professors. After all, breakthroughs can come from unlikely places, as recently evidenced by the Argentine car machine who developed a new method for easing childbirth or the teenager who invented a promising test for pancreatic cancer.
The recent moves were especially notable because they involved large for-profit businesses. The hope is the decisions may pressure other companies to follow suit.
“There’s momentum among researchers, universities and funding agencies,” said Peter Suber, director of the Harvard Open Access Project, when asked about the Microsoft decision. “But there’s not a lot of momentum among businesses, so it’s very welcome news.”
For hundreds of years, researchers and peer reviewers gave away their labor in exchange for the prestige that comes with publishing in a top journal and getting their work out to the world. Those promises allowed for-profit publishers to charge increasingly high fees for subscriptions or access to papers.
But open science proponents argue the work belongs in the public domain, particularly since it’s often funded by government or nonprofits. The cheaper and easier it is for students, patients, doctors and researchers around the world to review the literature, the more likely it becomes to make connections that lead to bigger breakthroughs.
“Scientific knowledge is a global good,” said Richard Price, CEO of Academia.edu, a fast-growing open science forum that has raised nearly $18 million from Khosla Ventures and others. “It’s important that everyone has access.”
But there are risks and challenges too.
Not surprisingly, some traditional scientific publishers are fighting the shift, which undermines their business model — if not their very reason for being. The industry has lobbied for laws that would protect the status quo and late last year publishing giant Elsevier reportedly issued thousands of takedown notices to authors posting their works on alternative forums.
Separately, while the Internet easily replaces the distribution function of traditional journals, it’s a trickier problem to replicate the systems for peer review and prestige.
The moves by Microsoft and Johnson & Johnson followed the University of California’s decision last summer to establish a similar open access policy across its 10 campuses, joining Harvard, MIT and nearly 200 other universities. (Though the policies often include important exceptions — and some publishers have been more accommodating of these shifts than others.)
Last February, the White House Office of Science and Technology Policy released a policy memorandum directing federal agencies with more than $100 million in research and development expenditures “to develop plans to make the published results of federally funded research freely avaoilable to the public within one year of publication.”
Meanwhile, a large number of open access journals have emerged in recent years, including some created by traditional publishers. Companies like Academia.edu and ResearchGate have taken a different approach, offering a sort of social network for researchers to share, read and comment upon works (which may or may not have been previously peer reviewed).
Google has also taken steps in this direction. The Mountain View search company reposts much of its published research in its own open online forum, “encourages” publication in alternative venues and pays open access fees to ensure that certain works are accessible to everyone.
“When it comes to scientific research, we have consistently said that open access to publications speeds up research, accelerates innovation, and helps grow the global economy,” the company said in a May blog post.
“It’s going to accelerate the pace of research.” — Jim Pinkelman, senior director at Microsoft Research Connections
But one of the critical functions of traditional scientific journals has been to vet research through the peer review process, ideally filtering out papers where the science wasn’t replicable or rigorous (not that it always succeeds).
So open science repositories that treat all research as equal and lack qualified scrutiny can be a dangerous thing, particularly in a nation where almost half of citizens believe in creationism and nearly three-out-of-10 don’t think there’s solid evidence of climate change.
There are, however, institutional forces working against that, Suber said. Publish or perish remains the rule in academia, with papers in reputable journals a key consideration in earning tenure. As such, most serious academics have little interest in putting their work out on unknown forums without review. Most simply want to repost articles that have already appeared in respected journals so that more people can see them.
Even among open access journals, a sharp divergence in quality is already becoming apparent. The Public Library for Science, BioMed Central and eLife have emerged among the serious brands in the space, offering their own prestige. Many others have been labeled predatory for charging researchers for the privilege of publication, among other questionable practices.
“Open science doesn’t have to be different in its methods, it can be just as rigorous and reliable,” Suber said.
But others are exploring alternative approaches to conferring scientific status.
Academia.edu, which now boasts 7 million academics and 2.3 million papers on the site, is developing new social tools for surfacing the best papers and feedback, akin to the “upvoting” and “downvoting” on sites like Reddit.
“Any peer review system has to have a way to filter the signal from the noise,” Price said.
But for Microsoft’s part, its decision last week was driven by the desire to put traditionally reviewed research in front of more people.
“We think more open access to research findings is the right way forward,” said Jim Pinkelman, senior director at Microsoft Research Connections, in an interview. “It’s going to accelerate the pace of research. Institutions realize that, authors realize that and we realize that.”