Publication fees
Previously, I mentioned that there are some problems with the current model of scholarly publishing. What are they? Publication fees is the first thing that comes to mind.
For readers outside the field, here's roughly how it works. A team of scientists gets funding to perform a study. In the United States, two of the biggest funders are the National Institutes of Health and the National Science Foundation. The scientists do the work. After the work is done, they write up a manuscript.
They submit the manuscript to a journal. An unpaid editor, essentially a volunteer at the journal, decides whether the manuscript is a good fit, and if so, contacts several reviewers to critique the manuscript. The reviewers are also unpaid volunteers. The manuscript goes back and forth between the scientists and reviewers until the editor finally decides it's ready to be published.
Next, journal staff format the manuscript into an article and upload it to their website. In my experience, this step can be rather lackluster, with journal staff missing critical typos or even introducing errors themselves. Mostly, articles end up as nicely formatted PDFs with pretty fonts, figures, and tables.
Publishers of journals are for-profit businesses. In the old days, publishers made most of their money by charging readers to access articles via expensive subscription plans to groups of journals, paid for by institutions like universities — in other words, the employers of the scientists who wrote the articles in the first place. An individual could buy access to an individual article, but this was often prohibitively expensive.
Perhaps you can see immediately why this was a problem. Publishers were the gatekeepers of knowledge. If you wanted to learn something, you had to pay them. Often this knowledge could be useful to people without the financial support of a research institution, such as the general public.
Enter the Open Access Publishing (OAP) model. The key difference between OAP and the previous model is that journals put articles on their websites for free without paywalls. This is unarguably the right thing to do. Now anyone can access knowledge without financial restrictions.
But there's a problem: how do publishers make money under OAP? The solution is to charge scientists to publish in the journal.
There are a couple of issues with this solution, the most obvious being the principle of the thing. Scientists do all the hard work, generate value for the readership in the form of knowledge, and generate content for the journals. And now they also have to pay for it themselves!
OK, this is an oversimplification. Scientists typically aren't writing personal cheques to journal publishers. The money is, or at least should be, budgeted within the original grants that paid for the work. (This is often easier said than done, as discussed later). Not every journal has moved to OAP; in fact, many journals now employ a hybrid model where you can publish an article more cheaply or for free with closed access, but pay extra for open access. Finally, some funders like the NIH have worked out a deal so that articles describing work they funded get uploaded somewhere else like PubMed Central without the scientists needing to pay.
It's still galling though. Imagine if Cormac McCarthy (or your own favourite author) had to pay Random House for publishing Blood Meridian (or your own favourite mid-eighties literary anti-Western novel).
The second issue is that we aren't talking about a flat rate, nor small amounts of money. OAP charges vary widely across fields and journals, but I usually see fees on the order of $3000. What's more, there's a correlation between price and journal prestige. Nature, one of the top journals in the world, currently charges about $12000 per open-access article. Papers in prestige journals get more attention both inside and outside of science. So if you want your work to be noticed— which you do, obviously — you pay more.
The third issue is that it disproportionately effects young investigators and small laboratories. These guys typically have fewer big grants and more small grants. A $3000 publishing fee is a big proportion of a $50000 seed grant and leaves less money to do the important stuff, such as paying research staff and buying equipment. By comparison, $3000 is almost trivial in the context of a $2.5 million NIH R01. In other words, paying for publications may mean that scientists get less done overall, and that junior scientists get less done than senior ones.
To be clear, I think the blame lies in the way for-profit publishers make money, not the OAP model. OAP is a wonderful idea. Many journals charge for closed-access publication as well, although they usually charge much more for open access, which really makes no sense because the extra work required to not put an article behind a paywall is trivial.
What can be done? Preprint services are a partial solution. A preprint is a version of a manuscript that hasn't been submitted to a journal yet. Scientists can upload their preprint to a service, such as bioRxiv , before publication in a journal. This means that the work can be read and cited by others while the scientists scrounge around for the money to cover the publication fees or even publish for free in a closed-access journal.
However, preprint services bypass the most important aspect of publishing in a journal: peer review. The extent to which the peer-review process actually works as intended is another issue entirely, but generally speaking, peer review is an essential part of the scientific process. The preprint system relies on the assumption that work will be peer-reviewed elsewhere — otherwise, anyone could write any old rubbish and upload it to a preprint service. Scientists still need to publish conventionally, and therefore pay for the privilege of doing so.