November 12, 2013 | 3
The first scientific journals were published in the late 17th century, and these print publications changed very little over time. Developments in printing technology, distribution and the advent of the commercial publisher all impacted the process, but the basic form was easily recognizable.
Then, the internet was born. Scientists and publishers hailed the World Wide Web as a force for transformation in scientific communication, and many journals started making articles available online. But little changed. The traditions of print publication were strong, and publishers created websites that mimicked the structure of the print publications.
Most current journals still have some vestiges of their print origins (even if they no longer publish a print edition), and many “born online” journals try to mimic their formerly-print competition.
In the days of print, lots of people spent a lot of time thinking about how to make journals smaller (fewer pages, and therefore cheaper to print and ship) while still being readable. As a result, many journals developed citation styles that were extremely abbreviated, often leaving out the article title or abbreviating the journal title.
Online, there is no need to count characters in order to save space, yet many journals still use these abbreviated citation styles. In 2013, the only thing these abbreviated journal titles do is confuse non-experts and force them to spend more time Googling things. Seriously, Int. Congr. Ser.? Rev. Roum. Chim.?
Many journals also choose to omit article titles from their abbreviate citations, a practice which frustrates me. While readers looking at the text can usually tell something about the article by the way it is used in the text, sometimes this is incredibly vague (or even misleading). Complete article titles are just helpful to readers trying to understand the provenance of the article in hand.
Publishers, please, save us all some time looking things up by simply spelling out article and journal titles. Make sure you provide DOIs to make things easier to find, and turn those DOIs into links to save us time.
For those few journals still publishing a print edition, page numbers are necessary, certainly. But fewer and fewer journals still publish a print edition (and fewer and fewer libraries subscribe to them). For online journals, page numbers introduce an artificial ordering of papers that often just seems silly, since most articles won’t be viewed alongside the articles that come before and after them. So why do online-only journals still use page numbers? It is an artificial artifact of print publication, and article numbers (or just relying on DOIs to be the article identifier) would simplify the process of finding things.
In a print publication world, articles would be attached to a physical issue of a journal, published and distributed at a specific time. If an article was received too late, it wouldn’t make it in. With the advent of the web, there is no longer a need to publish a group of articles all at once. Articles can be published as they are ready.
Publishers that cling to this hold-over from a print world end up introducing confusing complications into the publication process. An article may be published online in June, but not be part of an official issue until January (even if no print issue is published). This “official issue” doesn’t actually change the availability of the article. This actually creates some opportunity for publishers to manipulate their impact factors (and other metrics) as they make an article available months before its official publication date, allowing researchers more time to use and cite the publication.
Other articles may be needlessly delayed as editors wait for other articles in the “issue” to get through copyediting and typesetting. Topical collections of articles may still be handy, but they don’t need to be issued all at the same time. I’d like to see more journals drop the artificial artifice of the “issue” in favor of continuous publication.
As journals went online, the easiest way to transition was simply to post an online copy of the document that still looked like a print publication. The format for this is the PDF. On one hand, the PDF prints out nicely for reading on airplanes during takeoff (oh, wait, never mind). But by relying on PDFs, publishers lost opportunities to take advantage of the tools HTML offered. Simple advancements such as linking within the document to article sections or citations or linking to other documents took ages to become the norm, and some born online publications still don’t offer such tools. I still come across publications that post entire issues as a single PDF, adding additional steps to find the information you need. Although PDFs are nice to print, they place limits on what can be done with the article content. Tables can’t be downloaded and data can’t be re-analyzed. High quality images can’t easily be saved. And it takes longer to look up entries from the works cited section.
Thankfully, more and more publishers are experimenting with HTML or XML or even EPUB versions of journal articles. New features include larger print, internal navigation, linked citations, and downloadable images. Interactive features are becoming more common. The giant publisher Elsevier is working to implement so-called executable papers, in which readers can run code while reading an article. The journal F1000Research just announced a new data analysis tool built into their articles (published prior to peer review) to enable reviewers and readers to manipulate and re-analyze data. Hopefully new and existing publishers will continue to experiment with journal article layouts and features in an attempt to take advantage of the opportunities found on the web.
The Peer Review Process
The blind peer review process as we know it was developed in the second half of the 20th century as editors no longer had the time or knowledge to critically review all submitted manuscripts. Manuscripts were sent (by mail!) to the editor who sent copies to a few reviewers (by mail) who sent comments back to the editor (by mail). It was impossible to solicit the opinions of too many people prior to publication, and it was difficult to ascertain if readers thought the study was well done or interesting after publication. Since space was at a premium, editors needed to be selective, often rejecting solid scientific articles because they weren’t important or unique enough.
As journals moved online, the new online peer review systems were built to mimic this process, minus the photocopying. But the timelines remained the same – it still takes ages for a manuscript to go through peer review at most journals – and editors still routinely only seek the advice of a few experts prior to publication.
The instantaneous communication made possible by the internet offers opportunities for experimenting with the peer review process, but few journals are even considering changes. Making change is definitely difficult. The current system is tied closely to academic reward systems, entrenched beliefs about the quality of blind peer review (something that is very difficult to study, see Couzin-Frankel, 2013) and the lack of financial incentive encouraging publishers to innovate.
There are a new batch of journals born online that are taking advantage of the opportunities afforded by online communications, and several established journals have run pilot programs testing changes in the traditional process. Some journals (like the PLOS journals, and many BioMedCentral journals) allow registered users to comment on portions of an article or ask questions of the authors. Other journals (like PeerJ and F1000Research) are experimenting with posting reviewer comments alongside the published article.
Although there is some dissatisfaction with the current peer review system, most researchers are happy with it (Mulligan et al., 2013). Academics are a rather conservative lot when it comes to changes in their professional lives, and many see no reason to change. But most academics would agree that traditional blind peer review is imperfect, and the internet allows opportunities to explore methods of improving this process.
I look forward to seeing more journals experiment with open peer review, post-publication peer review, user comments, and specialty reviewers (e.g. statistical reviewers). These experiments and new strategies help us teach students about how the scientific process works, as well as igniting conversations among scientists about the strengths and weaknesses of the publication process.
What am I missing? What other aspects of the scholarly publication system are simply a holdover from a print publishing world? Is the article itself no longer necessary?
Couzin-Frankel, J. (2013). Secretive and Subjective, Peer Review Proves Resistant to Study. Science, 341(September), 1331. doi:10.1126/science.341.6152.1331
Mulligan, A., Hall, L., & Raphael, E. (2013). Peer review in a changing world: An international study measuring the attitudes of researchers. Journal of the American Society for Information Science and Technology, 64(1), 132–161. doi:10.1002/asi.22798