Jüri Allik, Professor of Experimental Psychology at the University of Tartu, belongs to the top one per cent of the world’s most cited scientists in his field. His recipes for becoming a top researcher were among the top 10 most popular posts on our blog in 2013. This is the last post in Professor Allik’s three-part series on scientific publishing. Read also the first post, Brave New World of Scientific Publishing, and the second post, The Crazy World of Peer Review.
Open Access
Scientists would be downright crazy if they silently accepted the kind of situation detailed in my first post. After just a little thought, the model of scientific publishing where the reader pays for the costs of publishing seems abnormal. It’s clear that new knowledge is mainly beneficial to the public. As most science is done with the taxpayer’s money, it would be expected that all scientific results achieved with such support must be publicly available.
The US Congress took it seriously and the Federal Research Public Access Act was proposed already in 2006, ordering all the publications made using federal money to be publicly accessible. What does this mean? It means that US scientists can no longer publish in journals that request money from readers in exchange for access to the contents. Right away, publishers smelled the possibility of further profits and made a proposal for those impacted by the act to make their articles publicly accessible by compensating not only the publishing costs but the owners’ sturdy profit, too. Because of that, in a strange way, some articles in the electronic versions of closed journals bear the tag of ‘public access’.
In the beginning of 2000, Nobel Prize winner Harold Varmus, Stanford biochemist Patrick Brown, and the computer scientist Michael Eisen posted a petition in which they called upon all scientists to boycott science journals with closed access. As I previously described, journals with closed access are those which request payment in order to read the entire year’s worth of issues, a single issue, or a single article.
At that moment, the great majority of scientific journals were closed. A reader who hadn’t paid for access could only read the title and abstract. Acknowledging the publishers’ right to profit, the authors of the petitions still made a compromise, allowing a variation of open access, which would come into effect after a moratorium of, say, six months. This should have been acceptable to all publishers, as an axiom of the publishing business is that most of the profits are made in the first couple of months after publication.
Brown and Eisen followed their call to arms with a project they named the Public Library of Science (PLoS). In 2002, philanthropists Gordon and Betty Moore gave nine million dollars to the PLoS as a gift. With this starting money, a series of journals launched, all with PLoS in their name (PLoS Biology, PLoS Medicine, etc). But the flagship would be PLoS One, which, at least at the level of declarations, does not discriminate against any field of science.
In the case that a manuscript is technically sound — it doesn’t claim the invention of Perpetuum Mobile or the positive impact angels have on alleviating the symptoms of depression — anyone can publish his or her work in the journal after paying a publication fee of 1,350 US dollars. (In real life, things aren’t that easy. In practice, the journal doesn’t exactly follow its promises. The scientists who are asked to review an article often act just as though it were the most typical closed journal. Still, one thing is certain: The shallow nature of quality control has had no dramatic effect when it comes to the quality of published works.) Of course, publishing involves an electronic version. Paper probably couldn’t handle the popularity of the journal.
In 2013, PLoS One published over 31,000 articles which made — it’s easy to calculate — over 40 million for the journal. 64 of those articles had at least one scientist working in Estonia among the authors.
Currently, PLoS One is the world’s largest scientific journal. In spite of the vast number of articles and low threshold of submission, the impact factor of the journal is quite high, at 3.7 in 2012 (This is the quotient between the number of articles published in PloS One in 2010 and 2011 and the references made to them in 2012).
Unfortunately, the bad habits of closed scientific publishing reached the open journals as well. One thing is clear: In the combination of a single fixed price for an article and the unexpectedly great torrent of articles, the profits have been growing at a quicker rate than the maintenance costs. The public was shocked when it turned out that Joseph Esposito, CEO of the PLoS journals, had a salary of over half a million dollars.
The number of open access publications has grown really fast. For example, in 2011 there was public access to half of the scientific articles published in the world. One could have thought that the world of closed publications, reigned by Elsevier, Springer, Wiley-Blackwell and other publishing houses, would just wither and die out. But the actual picture is quite the opposite.
The big publishers have gained the most profit from the open access revolution. As the scientists of many countries can no longer publish in closed journals, the payments have doubled. The library pays to subscribe to the journal, while the author pays to make his or her article public. Springer has been especially aggressive, as it bought one of the biggest open access publishers, BioMed Central, with all of its 258 journals, which continue as open access journals under the new owner as well. Actually, it doesn’t make much difference for the publisher if the author or reader of a scientific article pays for them. Probably the scheme in which the author pays has even less risk, as there will be no trouble distributing the journals.
The worst of all possible worlds
I hope that I have succeeded in showing that the existing world of management for science is by no means the best possible one. In some aspects, it’s even the worst, as it does practically nothing to protect the author’s dignity, let alone the truth. I can imagine the joy that everybody vocally criticising all kinds of bibliometry feels while reading this. One cannot trust the number of publications because the acceptance or rejection of an article is often biased towards some kind of ideology (of science). How can one keep a straight face when talking about the number of references if one can gain over a thousand references with data that’s basically fabricated?
In my opinion, it’s not very important how scientific significance is detected. For example, The John Hopkins University is much smaller than even the University of Tartu, be it the number of students or lecturers (the latter being 609). In spite of this, twenty researchers or professors from this university have won the Nobel prize over time, not to mention the eight Nobelists who began working there after they had already received their prizes.
Appreciation is definitely a good measure to use when evaluating a scientist’s scientific significance. One of Estonia’s science administrators has justified the obvious poor financing of Estonian science, stating that our scientists don’t deserve greater backing as they haven’t achieved anything remarkable thus far. There haven’t even been rumors about anyone chosen to be a candidate for the Nobel.
In my opinion, it doesn’t much matter if scientific significance is thought to based on the number of publications, patents, or, for example, grants received. Certainly, counting references has its considerable shortcomings, too, as it’s not the best indicator of whether somebody has had something important to say to the world. However, my mind refuses to accept knowledge of a couple of exotic details as scientific intelligence (e.g., something scandalous about the relatively unknown laundry bills of D.H. Lawrence), something which is nice to share with your two friends.
Where is the world of science heading?
It’s quite obvious that the system of closed scientific journals is being substituted for open access. It probably doesn’t even require green thinking to end the journals printed on paper, or rather, printing becomes a privatised affair – everyone can use on-demand printing to have a paper copy of a journal if he or she should happen to need it. One would doubt that peer review will remain unchanged as well, as two primary reasons for it are becoming extinct – the reader who finances the printing of a scientific article and the constraints on a journal’s volume.
One possibility is that peer review (censorship, in fact) will vanish altogether. Physicians, for one, have for more than ten years used arXiv, managed by the Cornell University Library, where one can post his or her article without it going through the ordeal of review. The format of an article must be .pdf or LaTeX. The archive can be freely accessed by all, and every month 7,000 manuscripts on average are added – quite an amount to read.
Another possibility is to make the review system interactive, offering the author a chance to argue with the reviewer if the former considers the latter to be wrong or unjust. Such a system is used by the journals published as Frontiers, founded by Swiss scientists Henry and Kamila Markramid. The first journal in the series was Frontiers in Neuroscience, and the first issue was published in 2008. In 2013, Nature Publishing Group bought all the Frontiers journals. Frontiers is distinguished by its interactive reviewing system. After an article is accepted, the names of the reviewers are made public, so, in a way, they share the responsibility for the article. Publishing an article there is not exactly cheap, but a scientist working in an establishment and getting grants can afford the expense of €770 to €1,600.
(A good alternative would be Sage Open, which asks only $99 for publishing an article shorter than 10,000 words. Although the journal is open to all social and humanitarian sciences, it hasn’t achieved popularity comparable to PloS One. This will probably change when WoS starts to index it.)
Many journals with open access have maintained the anonymous reviewers but lowered the threshold substantially. The PloS journals, for example, say that when reviewing, mostly the reliability and technical quality of the means used is evaluated, not the prospect of possible scientific contribution. Unfortunately, the reviewers often cannot follow the guidelines, as the line between the content and technical quality is often really vague.
When trying to conclude what is happening, one has to admit that open access hasn’t led the world’s science astray. Scientific articles produced under conditions of open access haven’t turned out to be much worse than the closed science with its strict censorship. Based on what has happened, one can predict that not very much would change if most of these 80 per cent of articles that the journals used to reject would in fact be published.
The only risk the journals would have is contributing to the already vast amount of articles that no one never references. On the other hand, it’s very likely that some of them would have a lot of citations and advance the science.
One of the virtues of open access publications is the modern technological platform inherent to them. PloS journals will tell the reader how many times the article has been accessed via Internet, how many times it has been downloaded as a text document, how many and which other publications have referenced it, and if it has made any waves in social media.
For instance, in 2009, the article on the “Genetic Structure of Europeans” by Andres Metspalu’s research group was published in PloS One. In the following 57 months, the article has been viewed over 12,000 times and the full article has been downloaded over 3000 times. In addition, we found that journals indexed by the Scopus database have already referenced the article 99 times and Wikipedia articles have mentioned it 14 times. So, the indicator of an article’s relevance would be its ‘life’ after it’s published and the activity surrounding it, which, in the case of the article mentioned above, has been remarkable.
With this in mind, it doesn’t seem strange at all that the four main authors of the article – Mari Nelis, Tõnu Esko, Reedik Mägi, and Andres Metspalu — are among the 1 per cent of scientists referenced the most in the fields of molecular biology and genetics, according to the current chart of the world’s scientists, renewed by the Essential Science Indicators every two months.
The Internet-based, open-for-all publishing has brought along the predictable extinction of many old conceptions. I remember the uplifting feeling and the smell of chalky paper from my distant youth, as I took the fresh issue of a journal out of the envelope, containing my own article next to others’. Surely now, approximately 200 articles later, the feelings are much more subdued, but the main thing is that the conception of ‘an issue’ is rather up to debate when it comes to electronic publishing.
Most standards of referencing, nowadays entrusted to computer programs such as EndNote, don’t even require including the issue number of a journal. As the pages are counted throughout the volume or the year, the numbers of these are enough for precise identification of the article. Although some electronic journals are still imitating a virtual single issue of a journal, there isn’t much point.
One of the best things about electronic publishing is that one doesn’t have to wait a year or so between the acceptance of a manuscript and its publication. In fact, it only takes a few minutes or an hour until the article is available on the web, after the decision to accept it has been made. The biggest delay comes from waiting for the publication fee to be transferred.
The numeration throughout the volume and the year is a thing of the past as well. A scientific article published on the Internet is an autonomous entity. Its location is no longer determined by these numbers. All of the more recent scientific articles have their own digital object identifiers (DOI). This is enough for identifying every article, as well as recognising and finding its location, or at least the address of its location.
All these developments hint that after some time passes, the concept of a journal as such will become vague, too. When there’s no peer censorship – I mean, peer review — anyone who has finished his or her manuscript can put it on his or her homepage where it will be available to the whole world.
Really, why should I pay €1,600 to someone else for putting my article on the Internet? At the moment, the trouble is the vastness of the Internet, with my article possibly just vanishing over there. But it’s highly likely that soon there will be a device that finds the articles, indexes them and makes them visible to all who are interested. Larry Page and Sergey Brin, the founders of Google, have already given scientists a present, as they have created a search engine called Google Scholar.
This search engine finds virtually every scientific publication on the Internet and identifies all other publications that have referenced it. Moreover, each scientist can create his or her researcher profile that will automatically start counting everything the scientist has published, as well as which other publications have referenced his or her output.
My good colleague Andres Metspalu, for example, has no trouble writing his CV, as it’s always possible to visit his Google Scholar profile and learn that by today (April 23, 2014) his works have been referenced 9,390 times. If no one else will do the job first, then Google alone could possibly make centralised journals extinct, with something that one could call the social network of science taking their place.
Still, I’d argue that scientific journals will not vanish altogether. The centralised editing practice, making enormous profits for the owners of publishing houses, will cease to exist. With distributed editing, all authors are editors as well. Naturally, there will be no use in maintaining the entire bureaucratic machine of the editorial staff. Anyone interested in this topic can just visit the e! Science News portal.
Compared to other analogous portals, it’s definitely not worse. It might even be better and more interesting, and, what is most important, more operative. The only substantial difference would be the lack of an editor. This electronic news engine has been directed by an artificial intellect since it was started in 2008. The code was written by Michael Imbeault, a student of virology and bioinformatics, using the high-level programming language Drupal, created by another doctoral student, Dries Buytaert. In my opinion, this is a really convincing example that there’s no point for a human being to do the job that a computer can do much better.
As I see it, the revolution looming in the near future has a sole loser – the big publishing houses who will lose their enormous profits. But when thinking about the great mass of scientists which has to – in addition to their main job and often instead of it – do the work of the editor and reviewer, which is not only mostly thankless, but also, as I hopefully made clear, simply meaningless, then the number of winners is bound to be much greater.
The biggest winner would be science, as the amount of pointless work would decrease. I really hope that some influential group of researchers will soon join forces with some capable programmer, resulting in a model of scientific journal using, for example, the above-mentioned Drupal, where there will be no editor or peer review. Registering could just require the address of a university, just as when creating a Google Scholar profile.
The quality of articles published would be determined by post-review: The number of all views, downloads, likes, comments, and references would offer a suitably diverse evaluation of the ideas in the article.
This is the last post in Professor Jüri Allik’s three-part series on scientific publishing. Read the first post, Brave New World of Scientific Publishing, and the second post, The Crazy World of Peer Review.