: Who Will Publish Research?

Who Will Publish Research? Open Access and the Future of the Scientific Paper

by Chris Hurst

There are strong reasons for librarians to be pessimistic about the future of scholarly scientific publishing. Journal subscriptions are becoming more expensive; budgets are always tight. I am not the first librarian to think that something must change, and so let me present one significant kind of change: "open access".

"Open Access" (http://www.arl.org/scomm/open_access/framing.html) means the free and unrestricted availability of scientific and scholarly literature on the internet. The term was adopted as part of the Budapest Open Access Initiative (http://www.soros.org/openaccess/) at a meeting convened by the Open Society Institute in Budapest in December 2001. "Open Access", as a term, resonates with "Open Source", and both movements share many features, not the least of which is the fervour of their true believers. Alas, I am not a true believer in anything, but even a stick-in-the-mud like me can see the potential for Open Access to change how the scholarly scientific paper is disseminated.

Any web surfer knows that there are enormous amounts of free text on the internet. Web magazines, blogs, chat rooms, and email lists are just some of the ways to read for free on your computer screen. However, the open access movement is not concerned with these. The movement is not concerned with any publication where the authors are paid directly for their contribution.

The challenges and opportunities of the open access movement are more important to scientific publishing than social sciences or humanities publishing.[1] Very little scientific scholarly publishing is done through books. While papers are natural candidates to be distributed through the web books are not. Should there ever be an electronic device that is as comfortable to read and as portable as a book, this may change. [2] For the moment, the paper, especially the scientific scholarly paper, is the focus of the open access movement.

The focus of the open access movement is on the papers that scientists and scholars produce and subsequently give away to publishers for the privilege of being published. The publishers take these papers, put them together in journals and sell those journals to libraries. In the case of electronic journals, publishers sell access to the journals.[3] In a sense, a university is paying for the research twice -- once for the research to be conducted and once to have the final product research accessible to the university community. The open access movement's ultimate aim is to transform the second cost, the cost of making the research accessible, from a cost that is paid by many universities and research institutions through subscriptions to a publishing cost.[4]

Of course, researchers do not give away the fruits of their research and expect nothing in return. The payment for this "given away" research comes in the form of tenure, promotion, increased access to grants. The researcher's goal in publishing is to give away their research to the most prestigious journal that will publish it, because the more prestigious the journal that publishes the research, the better the researcher's chances for future reward.[5] One of the challenges of the open access movement is to ensure that a researcher who makes papers available through open access is not penalized in terms of tenure, promotion, increased access to grants etc.

The cost of making the research available, i.e. the dissemination cost, has been one that libraries have traditionally paid. Every academic and research library has subscriptions to research journals. The problem is that commercial publishers own many of these journals and know that they have a captive market. Commercial publishers are beholden first to their owners, who demand that the publisher show a profit.[6] At some point, the commercial publishers realized that because they had a captive market, they could increase the cost of these journals at a rate that far exceeded inflation. When this was coupled with the ever-increasing number of journals, academic and research libraries were faced with a "serials crisis" (http://www.lib.uwaterloo.ca/society/crisis.html). The price increases for scientific journals have been especially steep. Expenditures for serials by research libraries increased 210% between 1986-2001, while the Consumer Price Index increased 62%. The typical library spent 3 times as much, but purchased 5% fewer titles. Libraries have had to cancel subscriptions, but libraries have also formed consortia to obtain better deals from publishers, and have turned to third party databases like EBSCOhost and used interlibrary loans to supplement the collection.[7]

The open access movement is only one response to the serials crisis. The Scholarly Publishing and Academic Resources Coalition (http://www.arl.org/sparc/) has been working towards solutions that will alleviate the serials crisis: open access, low cost journals that can directly compete with expensive journals and portals and other new outlets for scientific communication. The focus of this article is on open access, but this does not mean that it is the only solution.

No one disputes that there are costs in disseminating scholarly papers. Even if the internet obviates the costs of printing and mailing printed copies of a journal (by shifting the cost of printing onto the end user), there are still the costs associated with peer review and formatting the paper. The cost of formatting the paper can be transferred to the research institution(s), but since peer review has to occur at arm's length from the research being done the process of peer review cannot be handled by the research institution(s). Peer review is the one irreducible, non-transferable cost in the scientific publishing process.[8]


There are two main initiatives in the open access movement: self-archiving and open access journals. Self-archiving is the placing of a digital document in a publicly accessible website. These websites can either be subject-oriented websites (such as the physics preprint server at http://arXiv.org/ or the PubMed Central site at http://www.pubmedcentral.org/) or institutionally based sites.

The physics preprint (papers that have yet to undergo the peer review process are at the preprint stage) server at http://arXiv.org/ has been in operation since 1991 and is the great success story of the open access movement. Because the papers are at the preprint stage, the cost of peer review does not come into play. As of December 31, 2003, there have been 259,888 papers submitted to the server.

PubMed Central was originally intended to include preprint papers along with postpublication biomedical research papers (the original model for the site was the arXiv.org model), but the plan to include preprint papers was dropped in the face of opposition from many scientific societies. Instead, PubMed Central includes biomedical papers that have been published in recognized journals, but the papers are only added to the PubMed Central archive after an embargo period.[9]

The champion of institutionally based sites is Stevan Harnad. He argues (http://www.ecs.soton.ac.uk/~harnad/Tp/resolution.htm) that the most immediate and effective way to "free the entire refereed research literature" is for research institutions to establish Eprint Archives where all researchers at the institution can deposit versions of their papers (preprints, postprints etc.). There is open source software that will manage these archives so that the archives will remain "Open Archives Initiative" compliant.[10]

One of the key challenges of open archives is to ensure that all papers that have been placed in an open archive can be found by indexing software on the web. The Open Archives Initiative (http://www.openarchives.org) has published protocols so that each paper in each archive is consistently and accurately indexed.

A related challenge will be to include the open access archive URLs for the papers in indexes like EBSCOhost. Most indexes will include the URL for the paper on the publisher's web site where the access is restricted. If I follow a link in an index and find the access restricted, I am not inclined to do another search to see if that paper is available in an open access archive.[11] In essence, the freely available version of the paper is invisible to me unless I am prepared to do some extra work.

The various open archives projects are not in the business of peer review because they are conceived as supplements to the existing journals rather than as replacements. The papers in these archives are published, usually in a slightly different form, in journals. (Because the papers in the archives are slightly different, the authors can sign over the copyright to the published paper without having to remove the paper from the archive. However, some publishers will not accept a paper that has already appeared in an open access archive in any form. Researchers need to read the policy of any journal they are submitting papers to, and act accordingly.) Harnad argues that there will always be a role for publishers as "peer-review service providers," even if there is no call for their other services, like printing or distributing the journals.

Open Access Journals

Open access journals, on the other hand, are peer-reviewed journals that are freely available on the internet. This is accomplished by charging the researchers who successfully submit papers to the journal a fee. (Publication fees are not original to open access journals, as many journals have always had page charges or other fees associated with publication.) In some cases, the fee may be borne by the researcher's institution in the form of memberships in the organizations producing the journals.

The Public Library of Science (http://www.plos.org) journals ask $1500 (US) a paper, although they say "the ability of authors or their institutions to pay publication charges will never be a consideration in the decision of whether to publish." Discounts are available to institutions that pay for Public Library of Science memberships. BioMedCentral (http://www.biomedcentral.com) journals ask about $500 (US) a paper, but submissions from researchers at member institutions have the fee waived.

The Directory of Open Access Journals (http://www.doaj.org/) has just added its 700th record. This does not include general interest magazines or promotional newsletters that are freely available on the web. The directory also does not include any journal that makes its papers freely available after an "embargo" period (usually six months to two years). It is safe to say that there are enough open access journals being published to assure that the concept of open access journals is viable. However, there are still upwards of 20,000 journals that are not open access.

A Look into the Future

Will the open access movement alleviate the serials crisis? The short-term answer is no. The open access movement is still in its proving-the-worth-of-the-concept phase. I may think that open access is the best option for the future of scholarly scientific publishing; as a librarian I am predisposed to believe that. There are many people, both scholars and publishers, who are financially and emotionally invested in the traditional means of scholarly publishing. They disagree with me about the future and see no reason for wholesale changes at present.

One cause of the serials crisis that will have a direct impact on open access is the seemingly bottomless supply of scholarly papers waiting to be published.[12] In the short term, that means that even if some researchers choose to publish in open access journals, traditional journals will have no shortage of papers to publish. Traditional publishing acts as a limit on the number of papers being published simply because there is only so much you can print and distribute. Open access journals expand that limit (and add a further burden to the researchers and supporting institutions) without alleviating the subscription costs to libraries. The most cash-strapped research libraries could cancel subscriptions in the hope that open access journals, consortia purchases, third-party databases and interlibrary loans can make up for the loss, but no one will be happy about this.

The ultimate check on the number of papers published is the scholarly community's capacity for peer review. The capacity is not just dependent on the number and activity of reviewers but also on the capacity of publishers to administer the peer review. But why should the peer review process be tied to any traditional notion of publishing? A paper could be peer reviewed and placed in an open archive without being published in the traditional sense. The same agency that co-ordinates the peer review could make sure that the paper appears in the appropriate indexes. Even though the paper was in an open archive, it could carry an annotation saying that it was peer reviewed by an "X" agency on "Y" date, a claim that could easily be checked against "X" agency's records. In short, a paper could be available and peer reviewed without being published in the traditional sense, and what would be lost? The prestige of being published in "Z" journal, but in time, "X" agency could acquire the same kind of prestige as "Z" journal. "X" agency could be a scholarly society, a co-operative based in a discipline or a for-profit operation that has grown out of the existing publishing structure.

I do not know how scholarly scientific papers will be published/distributed in twenty years time. The internet has changed everything, and yet more journals are printed every year. Will open access be a large part of the scholarly process? Yes. Just as open source software will always be part of the computing landscape, open access will be part of scholarly communication. Are there challenges for open access to meet? Yes, but none of the challenges is insurmountable. Will it be the only form of distribution? Probably not. I have probably failed to account for some new technology that will, once again, change everything. The future is never quite what you think it will be. Having said all that, I think that the current model for scientific publishing has to change and open access is the best option for the future of scholarly scientific publishing.


  1. Scholarly scientific publishing is not a monolithic structure. There are many kinds of publisher: commercial and non-profit, big business and cottage industry. Each discipline has its own mixture of scholarly publishers. If some of my generalizations do not match the specifics of publishing in a discipline, I apologize.
  2. Do not hold your breath. I do not read anything longer than a few pages online. Electronic distribution shifts the cost of printing from the publisher to the reader.
  3. Some publishers will sell papers to third party databases like EBSCOHost who, in turn, aggregate many papers into databases and sell the "full-text" databases to libraries. Many publishers only provide the papers after a certain period of time has passed since the papers were originally published (an "embargo period"). The principle behind these embargo periods is that libraries will continue subscriptions if they are only getting delayed access to the papers through the databases. Convenience (for users) and value for money (for libraries) make these databases very attractive, even if their coverage of many subjects is hit-and-miss.
  4. Exactly who pays this publishing cost is an important question. Open access will not succeed if it is seen as a way for libraries to move the cost of disseminating from the libraries budget to the researcher's budget. To succeed, open access will have to involve co-operation between an institutions' libraries and researchers.
  5. There is always controversy about just how prestigious any journal is, but it is safe to say that some journals are more prestigious than others.
  6. I do not have an ideological problem with this, but I do not think library budgets should be trapped in the invisible hand if there are alternatives.
  7. Brandon University has done very well by being a member of the Canada wide consortium CNSLP (Canadian National Site License Program). We should continue to support this program as it represents a real opportunity for small Canadian universities.
  8. Of course, few people like the process of peer review. I would entertain arguments that the process has to be reformed, but I do think it is essential.
  9. For the complete story on the creation of PubMed Central, see the paper "The Real Stakes of Virtual Publishing: The Transformation of E-Biomed into PubMed Central" in Journal of the American Society for Information Science and Technology 55(2): 127-148 2004. - http://www3.interscience.wiley.com/cgi-bin/abstract/106563225/ABSTRACT
  10. Setting up and maintaining these archives are not trivial tasks, but, without examining the software in depth, I cannot tell you just how difficult the tasks are.
  11. Perhaps I am lazier than most people, but I doubt it. Active web surfers overestimate the average person's interest in spending time on the web.
  12. I do not have statistics for this. Measuring the number of unpublished papers is an enormous task, and I am not aware of anyone who has done this. However, there is ample anecdotal evidence that there is a glut of papers to be published. I would be happy to be proven wrong about this.

About Ecclectica | Current issue | Issue archive | Links | The editorial team | Contact us
ISSN 1708-721X