I’ve said before that for scientists it doesn’t matter what great experiments you are doing to uncover new truths about our universe if you don’t share that information with others. Scientific publications are the official way scientists communicate with one another and presumably the general public about their research progress. How does this process work and how well is it serving its purpose? Let’s take a closer look.
Practically speaking, here’s a review of how it works. Scientists must first decide which journal among many available is appropriate for this new work. Different journals have different foci and levels of prestige, and you can only submit your work to a single journal at a time. A manuscript is submitted to the journal along with a cover letter to the editor explaining its suitability for publication. If the editor agrees with its suitability, it is sent out to two or three other independent scientists for their reviews and recommendations to accept, revise, or reject the manuscript. The editor synthesizes this information and relates a decision to the corresponding author. The author(s) must then respond to the reviewers’ concerns with additional experiments, clearer writing in the manuscript and/or logical rebuttal in a separate letter responding to reviewers’ comments. If the revised manuscript is still not acceptable for publication, it can be sent to another journal to undergo the same peer-review process. If the revised manuscript is acceptable, then it proceeds through the publication pipeline where the article is further formatted for final publication. Usually accepted manuscripts are available online before they appear in print. Finally, the article appears in a certain volume and issue of the journal, officially part of our collective human knowledge.
Overall, this process is meant to improve the quality of scientific publications by ensuring that new, non-duplicative, high-quality work vetted by independent researchers is presented. On the surface, the system seems logical enough, but it does have some weaknesses.
Scientists can do a better job at writing up their work. While the audience for these types of publications is other scientists, publications are notorious for their use of jargon requiring all but the most expert of readers to track down multiple other publications from the reference list to understand the current report. At the other end of the spectrum, this medium functions poorly as an outlet for communicating science progress to the public.
There is a premium for articles describing something completely novel and interesting. It’s always great when the science comes together in that way, but these new stories are usually built upon a lot of decidedly un-flashy experimentation. This kind of science is still sound and deserves a place in print. This novelty bias also favors positive results over negative ones. Negative data can sometimes be valuable, but it is always difficult or impossible to publish. Usually the fate of such negative or orphan data (legitimate results that cannot be explained) is to languish in laboratory notebooks, far-removed from being useful for scientific progress. However, some negative data can be useful if for no other reason than it prevents future scientists from investing in similar experiments. Check out this post calling for the publication of negative data.
If two groups of scientists are working on similar projects and one group publishes their findings first, there is no consolation prize for second place. It is exceedingly difficult for the second group to get their results in print. The second-place group must have some aspect of novelty or push the results further than the first in order to be published. This seems reasonable, given that it is counterproductive and uninteresting to read about duplicative research. However, the pressures of these competitive fields can create an environment conducive to flawed or fraudulent research. In this context, having reproducible studies could be useful for verification. As it stands now, agreement is boring and thus unpublishable and disagreement is not allowed; the burden is on the second-to-publish to refute any conclusions of the first-to-publish. Note, that some science is so competitive and similarly paced that duplicative research between two groups can be submitted and reviewed by different journals leading to their appearance in print at roughly the same time.
Finally, there is the matter of access to scientific publications. Traditional scientific journals operate on a subscription basis for access to their articles. Students, faculty and staff have access via their institution’s subscription, but these can vary significantly among universities. Members of professional societies usually have access to the journal(s) sponsored by those groups. Everyone without access must pay for it. This paywall issue is another example of why traditional scientific publications are not a good medium for public science communication. Stay tuned for tomorrow’s post on a new breed of science publication- the open access model.
The beauty of science is that it is continually re-evaluating itself for optimization. The area of scientific publication is poised to undergo a significant systemic evolution as a result of the advent of the information age with new communication technology emerging at a rapid pace. The future challenges related to scientific publications will be increasing the speed at which reliable information is available, improving dissemination to a wider audience in ways that are understandable (without diluting the message too much), finding a place for orphan science experiments and identifying/flagging fraudulent or erroneous publications. Whatever form these new solutions take, must also be in a format stable enough to endure as a piece of accessible information for centuries to come.
Extra reading and resources:
How journals work (A slideshare presentation from an insider)