Stevan Harnad, Canada Research Chair, University of Quebec at Montreal
To remind ourselves how different is the author of a peer-reviewed journal article from every other kind of author, we just have to recall the purpose of universities' "publish or perish" policy: Aside from imparting existing knowledge to students through teaching, the work of a university scholar or scientist is devoted to creating new knowledge for other scholars and scientists to use, apply, and build upon, to the benefit of us all. Creating new knowledge is called "research," and its active use and application are called "research impact." Researchers are encouraged, indeed required, to publish their findings because that is the only way to make their research accessible to and usable by other researchers. It is the only way for research to generate further research. Not publishing it means no access to it by other researchers, and no access means no impact -- in which case one may as well not have done the research in the first place.
So it is the need for research impact that makes the author of a peer-reviewed journal article different from every other kind of author. The author of a book, textbook, or magazine article might even be a peer-reviewed research author wearing another hat, but the difference is like that between night and day: For the author of a book or textbook writes in order to have the text sold for royalty income, and the author of the magazine of newspaper article is writing a work for hire, for a fee or salary. Not so the researcher, who publishes in the peer-reviewed journal solely for maximal research impact, never seeking or receiving a penny from the sale of the text. On the contrary, researchers have traditionally, at their own expense, mailed reprints of their articles to anyone who requested them, so important was it to them that their research be read and used.
Why were researchers (and their universities) actually willing to pay to maximize the accessibility of their research output by disseminating reprints? Because access is a prerequisite for impact: Anything that blocks access blocks impact. The unread article is the unused, uncited article. This is also why citation-counts -- "how many papers have cited my paper?" -- have become such important performance indicators for research uptake and impact. The more a piece of research is used in further research, the more it has contributed to knowledge. And both the universities' publish-or-perish reward system (of salary, promotion, tenure, prizes) and the public and private research funding system (of grants to researchers and overheads to their universities) are based on measuring, predicting and rewarding research impact.
But something has changed. Researchers and their universities are beginning to realize that the online era has made it possible to enhance their research impact dramatically. It is no longer necessary to expend the effort and cost of mailing out individual reprints of one's peer-reviewed articles; they don't even need to be emailed any more. They can be publicly self-archived in the university's Eprint Archives -- websites that are accessible to all would-be users worldwide, without anyone having to make or respond to reprint requests: http://www.eprints.org/self-faq/
The transition began spontaneously: Researchers began to post their papers on their own websites, to be found by would-be users through google. But this was a bit like finding a needle in a haystack, unless the user happened to know in advance the title of the paper and the author. It was certainly no substitute for searching for the paper in a focussed database consisting of only peer-reviewed journal abstracts such as Medline. But Medline, with its focus, lacked the full-texts of the papers themselves, whereas google, with its universal reach, lacked the focus to find them.
The solution was twofold. First, The Open Archives Initiative (OAI) http://www.openarchives.org/ created a protocol for tagging the critical metadata identifying research articles (author, title, journal, date, abstract, keywords) so that all papers that were compliant with the OAI protocol would become "interoperable," meaning that they could be harvested, searched and retrieved as if they were all in one virtual archive containing all and only peer-reviewed research. The second step was to design (free) software that would create OAI-compliant university Eprint Archives -- http://www.arl.org/sparc/core/index.asp?page=g20#6 -- in which authors could immediately deposit all their articles so as to make them openly accessible to all researchers, thereby maximizing their impact. This spawned OAI harvesters such as OAIster http://oaister.umdl.umich.edu/o/oaister/ which now allow researchers to search the archives of 167 OAI-compliant institutions, already containing over a million records.
The infrastructure for maximizing university research impact is hence already available or in place. What are urgently needed now are tools and policies designed to create and fill the university Eprint Archives as soon as possible, for until those archives are filled, research impact is being needlessly lost every day.
(1) Universities need to adopt a self-archiving policy -- an extension of their existing "publish or perish" policy to "publish with maximal impact". A potential model for such a policy can be found at http://www.ecs.soton.ac.uk/~harnad/Temp/archpolnew.html along with (free) software for creating a standardized online university CV, linking all entries for peer-reviewed articles to their full text self-archived in the university eprint archives: http://paracite.eprints.org/cgi-bin/rae_front.cgi
(2) University libraries need to help with the first wave of self-archiving, doing "proxy" self-archiving for those researchers who feel too old, tired, or busy to do the few keystrokes per paper that are involved. http://www.ecs.soton.ac.uk/~harnad/Tp/resolution.htm#7.3
(3) Research funding agencies such as NSF and NIH need to encourage self-archiving as part of the research cycle, requiring not only that the research findings be published, as they already do, but that their visibility and usage be maximized by making them openly accessible. http://www.ariadne.ac.uk/issue35/harnad/
(4) Scientometric performance indicators and analyzers such as http://citebase.eprints.org/cgi-bin/search -- rather like google, but based on citation links rather than ordinary links -- need to be created and used to demonstrate, monitor and reward the maximization of research impact through open access. Free online accessibility increases citation impact by 336% http://www.neci.nec.com/~lawrence/papers/online-nature01/
(5) Journals need to support self-archiving by modifying their copyright transfer or licensing agreements to encourage self-archiving, as 55% of them already do, and most others will agree on a per-paper basis if asked (so ask!): http://www.lboro.ac.uk/departments/ls/disresearch/romeo/Romeo%Publisher%Policies.htm
There are at least 20,000 peer-reviewed journals, publishing at least 2,000,000 articles annually. Their impact could be at least 3 times as great if they were all self-archived. The financier George Soros's Open Society Institute's BOAI is now supporting open access http://www.soros.org/openaccess/ as is the Scholarly and Academic Resources Coalition http://www.arl.org/sparc/. The momentum of self-archiving is growing, but if universities and their research funders were to take the five steps outlined above in a concerted way, there is no reason why all their refereed research output could not be openly accessible, virtually overnight, for all other scholars and scientists to use, apply, and build upon, to the benefit of us all.