In my latest Pacific Standard column, I write about Nature Publishing Group’s new read-only access policy, allowing subscribers and select media outlets to share links that tunnel through the paywall. I argue that it’s time to get back to basics: We need to ask, why do we have science journals, and do we still need them in the 21st century?
Ever since their inception, science journals have served three primary roles:
#1 They disseminate research findings to the community
#2 They provide quality control by organizing peer review
#3 They serve as a record of priority and research accomplishment
In his Very Short Introduction volume on economics, the economist Partha Dasgupta has a nice explanation of how these functions of a journal were an important innovation of the Scientific Revolution, as a way to provide incentives for researchers to produce and share knowledge as a public good:
[The Scientific Revolution] created institutions that enabled the production, dissemination, and use of knowledge … to be transferred from small elites to the public at large.
Science journals are one of those institutions, but a lot has changed in the past 350 years. How well do journals meet our needs today?
On #1, you could argue that paywalls and snowballing costs of journal subscriptions inhibit the dissemination of knowledge. These days, it is so easy to distribute data and manuscripts. The cost of dissemination should be the cost of hosting the server and not much more. Journals with paywalls are therefore not the best available way to do function #1.
But, that’s not the only thing a journal does. Function #2 costs money, and therefore journals need revenue. How well are journals carrying out function #2? Maybe not so well. We are plagued with published research that is not reproducible. On top of that, a lot of the peer review process involves discovering whether a manuscript is ‘suitable’ for a given journal, even if the data and methods are sound. It’s a wasteful process that results in a perfectly fine manuscript going through multiple rounds of review at different journals as the authors try to get it into the one with the biggest impact factor.
Which brings us to function #3: journals play a crucial role in allocating career rewards in science, because the papers they publish serve as a scientist’s research track record. The problem is that journals are maximizing their utility by trying to become the most high-profile place to publish – this serves the interests of the journal more than those of the scientific community. And it distorts the career incentives of scientists, who, in order to get jobs and tenure, have to play the CNS lottery. The distorting effects of Impact Factor maximization is a failure of function #3.
Journals aren’t about to go away, but it’s clearly time to rethink how a 21st century journal should operate, and reconsider whether there is a better system to carry out these functions.
The Finch and Pea 