There has been some fuss in the press concerning a recent study published in the open access journal Plos One. As an aside, I will just mention that open access journals like PLOS (Public Library of Science) and the Biomedcentral series are most certainly the way of the future. It is getting harder and harder to justify the old system of scientific publishing. How does the old system work? Most of the content of the journals are provided by academics. We write up our research results, submit it to the journal, where it is sent out by editors (who are mostly unpaid academics) to be reviewed by 2 or more of our peers, who are also unpaid for their services. Should the manuscript be deemed worthy of publication, the journal will then take the electronic version of the manuscript and figures and put it together for publication after assessing “page charges” to the authors, which can be quite expensive.
So let’s examine the equation as it stands. On one side we have the publisher who puts out the journal and charges libraries and individuals for subscriptions, while acquiring most of it’s content for little or no money (and in fact charges the content providers for the privilege of publishing). Moreover, they also own the copyright on all of this content. On the other side, we have the acadmics who produce the content, review the content, pay for reprints, and pay for subscriptions.
The key part that I’m leaving out of the equation, of course, is that academics need to publish in order to get their work read and to gain academic advancement. Certain journals are known to be more prestigous than others, and so getting your papers published there is certainly preferable than in a lower tiered journal, both in terms of the readership and in the fight for promotion. So we are bascially paying for the prestige of publishing in a well-respected peer-revieweed journal.
But it doesn’t have to be this way. It is ridiculous to assess the success of a scientist based on where he or she publishes. The real measure of a scientists success is the impact that their science has had on the field, which can be judged by how often his or her papers are cited in the literature. As an example, there are large numbers of papers (perhaps most) that have been published in Science or Nature (the top two multidisciplinary journals) that have never been heavily cited and have had limited long-term impact on the field. Conversely, there are seminal papers that have significantly altered scientific endeavors that were published in other journals and that have been cited continuously for decades. With the advent of all the indexing services and the explosion of information technology, it is a trivial matter to assess a scientist’s impact on his or her field by simply measuring how often his or her work is cited. Heck, it may even be that Google page ranks can serve the same purpose. There is no need to rank journals.
So if there is no need to rank journals, then there is no reason why all papers shouldn’t be published in open access journals. These journals operate similarly to print-based journals (manuscripts are peer-reviewed), with the exception that the author maintains copyright of their material, and the papers are immediately available to everyone: No subscriptions. No reprints. Yes, they do require a page charge, but that is waived for any academic who submits from an institution that supports the open access journal, so for most of us, it is free.
The next logical step, then, is to take peer review out of the equation as well. With open access publishing, the entire scientific community will serve as peer reviewers. Isn’t is better to judge the quality of a manuscript by having the feedback of hundreds or thousands of scientists rather than 2 or 3 chosen by the editor (and often chosen from a list of 4-5 provided by the authors)? In the short term, the quality of the science can only be improved as authors respond to critiques (anonymous or otherwise) posted by their colleagues. Over the long term, the quality of the publication will be judged by the number of times it has been viewed and the number of citations.
There are obviously some issues that need to be dealt with for this model to work. It already works in relatively small fields (like certain disciplines of physics), but scaling up to large fields in biomedical research do present problems. But I have no doubt that this will be the model going forward—the only question is how soon do we get there.
Now where was I before I was interrupted? Oh yeah, the paper on tool use in rodents. I think I’ll talk about that another time.