Are we outsourcing our identities to the academic journal industrial complex?
The perils of "It's not science if it's not published"
Welcome back to the latest series on publications. A previous series, looking at leaving academia through the lens of Conscious Uncoupling, is not finished but on hiatus. Previous newsletters on publishing are here and here and here.
Your daily steps don’t count towards fitness unless they were tracked by a smartwatch. Your vacations might as well have never happened if they weren’t captured on Instagram.
But did you know that experiments only matter if they are published?
At least, this is what I was told. It was my first or second year as a faculty member when another assistant professor proclaimed, “If it’s not published, it’s not science.” I don’t remember the context in which this was said, but I do recall being shocked to my core. I didn’t—and still don’t—agree with this statement at all.
Of course, publishing is important. Yes, peer review, transparency, and sharing data are fundamental aspects of academic science. Science is a communal activity and no one should be deliberately holding onto data. We need results to be made public, so that we can build on existing knowledge. The current publishing system also provides a mechanism for peer review and reproducibility. Equally important is to share the results of publicly funded work.
But—the idea that publishing is essential to doing science has been bothering me for 15 years! As I am thinking about my divorce from academia, the culture of scientific publishing looms large, and this particular idea keeps coming up as a sore spot. So today, let’s dig into this discomfort a bit. Here are some of the ways in which I think “it’s not science until it’s published” harms science and scientists:
A lot of research isn’t published and that’s okay. Negative results, preliminary work, projects that failed, half-done work, projects that lost funding before they were completed—all are typically not published. But far from being “not science”, these kinds of dead ends or false starts are fundamental to the way that discovery science is done. Imagine telling your PhD student that they spent 5 years doing something that wasn’t science when working on a failed project!
It’s true that it might be nice if there were a way to make negative results and partially finished projects available to the community, but that’s not the way that publication works in the sciences right now (for the most part). Publication bias (the lower likelihood of publishing negative results) is well documented.
And—what about research done outside of the academic structure? Plenty of research is done in industry, and much of it is unpublished. Kids do experiments in the kitchen, but don’t publish their findings about the effect of mentos on Diet Coke. What about Victorian ladies studying and cataloging moss? People help catalog birds or butterflies in community science projects.
All of this is science, isn’t it?
At a fundamental level, “it’s not science if it’s not published” conflates the industry of academic science with scientific inquiry. Science is a systematic way of understanding the world. Let’s not mistake the capitalistic structure in which scientists operate for the work that they are doing inside (and outside) that structure.
Making publishing a requirement to validate what we do as scientists is essentially a form of gatekeeping. “Publishing” means paying a journal thousands of dollars to handle peer review and dissemination of your work. So, with the “it’s not science if it’s not published” philosophy, we limit who can do science to those who can pay. As a result, there are documented geographical biases in publication rates.
And—newsflash—publishing is not always a sign of properly performed research. Plenty of questionable results get published—invented figures, over-interpreted results, sketchy approaches. Predatory journals will publish anything for profit. So why make the process of publishing, one that is known to be flawed, the bar for doing science?
These are just my initial thoughts, but I’d love to know yours! Please leave a comment below.
Discussion Section
What do you think? Is it science if it’s not published?
I sometimes do wonder how the current push for paid open access publications is going to lead to a further inflation of the system of "famous lab produces famous researchers".
We often get solely/mostly judged based on our publications. (I wish it wouldn't be the case, but it often boils down to that.)
And high publishing costs quickly become restrictive to different labs. Only those (usually big) and well funded labs can afford to pay multiple "high impact" publication fees a year. Which in turn increases their likelihood of winning a grant, attracting new recruits, publishing "high impact "again etc. and becoming an even more "famous lab".
Meanwhile those smaller (newer?), not so well funded labs can no longer consistently shine through their work alone.
And when it comes to gaining a next position: Who will have the "better" cv, just due to the funding status of their previous employer?
I fear this will lead to people seeking out "famous" labs for their employment even more. Thereby essentially becoming an even stronger culture of gatekeeping by who is "famous" versus "who isn't".
Liz, when you say "published", it seems you mean "published in a peer-reviewed journal". It is a good phenom that more and more institutions no longer distinguish between a preprint and a peer-reviewed publication, and consider a preprint as published output.
Also, regarding Basti's comment above, I would like to push back. The costs for the effort that goes into a typical molecular biology paper from a Global North lab is something like 250,000$ for salaries with benefits (guesstimate of 3 years average for 1 doctoral/postdoctoral researcher, plus 3 years of 10% PI salary), 50,000$ in consumables etc., 150,000$ in overhead, so on the order of 450,000$. So 5,000$ in APCs (article processing charges) adds about 1% to the total cost. Also, please remember that the goal is of course that eventually there will be no more subscription costs, and that OA publishing will be at least cost neutral, if not leading to overall cost savings for academia.
In the interest of full disclosure: I'm Deputy Editor of an OA journal, and I'm remunerated for my work as an editor.
Cheers,
Detlef Weigel