The Promise of 'Open Access' Publishing

I'm pleased to welcome as our guest Peter Suber, a philosopher at Earlham College and open-access project director at Public Knowledge (www.publicknowledge.org), who is both enthusiastic about the open-access movement and up-to-date about the issues the new publishing method raises.

The call for open access among scholars, universities, libraries, and foundations is limited to peer-reviewed research articles and their preprints. The reason is simply that we scholars and scientists do not receive royalties for our journal articles, as you well know. We might write books and software for money, but we write journal articles for impact. We willingly, even eagerly, submit our articles to journals that do not pay us for them, and we've done this since the first scientific journals were launched in 1665.

So at least with respect to our journal articles, we are very differently situated from musicians and movie-makers whose livelihoods might be threatened by open access. On the contrary, our interest is to share our work with everyone who can make use of it. We want to reach the largest possible audience so that our results can be noticed, read, taken up, built upon, applied, used, and cited. Universities and laboratories share this interest insofar as they want their research output to be as visible and influential as possible. Foundations share this interest insofar as they want their funded research to be as useful as possible. Governments and citizens share this interest insofar as open access accelerates research and all the benefits of research, from medicines and technologies to environmental health, economic prosperity, and public safety.

That's the goal. There are many means to this goal and many ways to pay for them.

Here are some major definitions and endorsements of open access.

Budapest Open Access Initiative (February 14, 2002)
http://www.soros.org/openaccess/

Bethesda Statement on Open Access Publishing (June 20, 2003)
http://www.earlham.edu/~peters/fos/bethesda.htm

Berlin Declaration on Open Access to Knowledge in the Sciences and Humanities (October 22, 2003)
http://www.zim.mpg.de/openaccess-berlin/berlindeclaration.html

World Summit on the Information Society (December 12, 2003) Declaration of Principles and Plan of Action
http://www.itu.int/wsis/documents/doc_single-en-1161.asp
http://www.itu.int/wsis/documents/doc_single-en-1160.asp

Here's some further reading to put open access in perspective and help answer questions about it.

Landmarks in the evolution of open access
http://www.earlham.edu/~peters/fos/timeline.htm

The best FAQ on open access for people new to the concept, from the Budapest Open Access Initiative
http://www.earlham.edu/~peters/fos/boaifaq.htm

What can universities do to promote open access, with separate sections on faculty, librarians, administrators, and students
http://www.earlham.edu/~peters/fos/lists.htm#do

Publishers are divided on it, and even those who support the goal of OA wonder how it will affect them. This creates transition problems, I agree, but it is certainly not a tragedy of the commons. OA literature *prevents* classical forms of the tragedy of the commons. Freely available online digital literature is a resource that is not diminished by use, unlike grazing land or fishing grounds.

If you're saying that strict pursuit of self-interest will foreclose this public good, then I don't agree. It may be that some publishers see their private interest opposed to OA, but it is not true of all. Some see that the current subscription model is not sustainable. The University of California Academic Senate just called it "incontrovertibly unsustainable". Some publishers are experimenting with OA as way to ensure their own survival. But even apart from this, scholars can realize open access without persuading or converting publishers. We are the ones who write the articles and perform peer review. In the age of the internet, we can offer one another these services without the mediation of publishers who decide not to participate.

I also agree, by the way, that one way to overcome the transition problems is to change incentives. The private foundations that signed the Bethesda Statement on Open Access Publishing changed incentives by offering to pay the processing fees charged by OA journals, http://www.earlham.edu/~peters/fos/bethesda.htm. Many public funding agencies in Europe agreed to do the same when they signed the Berlin Declaration on Open Access to Knowledge, http://www.zim.mpg.de/openaccess-berlin/berlindeclaration.html. Universities can change incentives by bringing their promotion and tenure criteria into alignment with new developments in the distribution of scholarship, by creating OA institutional repositories, and by encouraging their faculty to deposit their research articles and preprints in them. I have some other ideas on how universities, foundations, and governments can create new incentives to realize this public good, http://www.earlham.edu/~peters/fos/lists.htm#do.

1. Where does the revenue for the long-term archiving, which could be costly, come from?

2. Where does the revenue for updating the formats, interfaces, searchabilities, and other technologies come from?

3. If libraries/information seekers need tech support and customer services, who should they turn to?

4. When you buy an expensive product, you care about the quality and complain/ask for a refund if the quality does not meet your expectations. You will certainly not ask for refund if your free lunch tastes awful. How do you make sure that the quality will not deteriate when pricing is no longer the supporting foundation of quality?

Thank you.

Peter Suber:
    Good question. There's no doubt that long-term preservation is as important to online literature as it is to print literature. Let me take BioMed Central as an example of an OA publisher that has taken effective steps to address this problem.

BMC publishes over 100 OA journals. First, it publishes a print edition at cost for constituents who want it. That gives BMC journals the same longevity as print journals, which is enough to erase the special doubt that is sometimes directed toward online literature. Second, BMC has written an "Open Access Charter" to insure the OA of its corpus in case the company fails or is bought out. Third, BMC deposits every article it publishes in PubMed Central, an OA archive hosted by the U.S. government. This gives very practical effect to the Charter by assuring another avenue of OA to the BMC literature regardless of the fate of BMC itself. Fourth, BMC has arranged for its corpus to be archived and preserved by the National Library of the Netherlands. It also lets any user download the entire corpus for text-mining, insuring the existence of multiple copies around the world. Finally, BMC marks up all its articles in XML, which greatly reduces the cost of future formatting changes required by new technologies.

Where does the money come from? As you can see, there are many answers. Some libraries pay for a print edition to its journals. Some governments pay for long-term preservation and access. BMC pays for many of these steps itself.

I should add that conventional electronic journals have such burdensome licensing restrictions that libraries are often not free to make copies, store them long-term, or transfer them to new media and formats in order to keep up with changing technology. By permitting all these forms of copying and adaptation, OA greatly increases the odds, and opportunities, for long-term preservation.

If you're saying that OA journals might accept weak papers, papers that would not ordinarily survive their peer review process, simply to collect a fee that only covers their costs, that's very far-fetched. First, similar conflicts of interest arise at conventional journals, e.g. when an author works for an advertiser. All experienced editors think about these conflicts are used to handling them. If anything, the problem is worse at subscription-based journals where profit margins are high and price hikes are publicly justified by the growing volume of articles.

Second, all peer-reviewed journals, regardless of their funding model, have strong professional reasons not to reduce their quality. It's only by maintaining their quality that they can attract readers and authors. The Executive Director of the Public Library of Science (PLoS) is Vivian Siegel, formerly the Senior Editor at Cell. She did not move from Cell to PLoS in order to lower her standards and would not let the fact that PLoS uses a new business model alter her scientific judgment. Moreover, every peer-reviewed article at PLoS is seen by at least one academic editor who does not work for PLoS, is not paid by PLoS, and who has no financial stake in the outcome.

If you're saying that OA journals might accept more *strong* papers than their conventional counterparts, and do so in order to collect the processing fees, this remains to be seen. If it happens, you could criticize it as increasing information overload or praise it as a valuable way to take advantage of the new medium. But you couldn't complain that it compromised peer review.

On the second half of your question. It's true that the upfront funding model works best in fields like biomedicine where most research is funded and most funders are on record as willing to pay the fees. In less well-funded fields, including the humanities, we have to explore other models. Many OA journals in the humanities do not charge any processing fees. For example, Philosopher's Imprint is edited by philosophers and published by librarians at the University of Michigan. Because both groups are already on the university payroll, the journal doesn't need to charge any fees, either at the author end or the reader end. Overlay journals are another way to reduce costs so that journals needn't charge processing fees. (Overlay journals use OA archives or repositories overlaid with a peer-review service.) I have more details and ideas here, http://www.earlham.edu/~peters/fos/newsletter/11-02-03.htm#objreply.

All universities should have an OA institutional archive or repository. The easiest way to fill it is for the promotion and tenure committee to encourage or require candidates to fill it. If the committee asks for a standardized vita with working links to all the OA articles, that might be enough. Or it could decide to consider only articles on deposit in the institutional archive. This does not deprive authors of the freedom to publish in the journals of their choice. It only enlarges the visibility and impact of the research, the author, and the institution.

An interesting variation could be called sponored OA. For example, Quigen recently sponored six months of free online access to a group of Nature articles on RNA interference. One of the most promising but least explored possibilities is endowment. There isn't just one way to cover the costs of a peer-reviewed journal, and we're still very far from having exhausted our cleverness and imagination.

I should also add that OA archiving does not require the revenue stream that OA journals do. OA archiving is a direct avenue to OA that does not require the funding, let alone the funding ingenuity, of OA journals.

How will their fair share of the overall cost of publishing be recouped in Open Access?

Peter Suber:
    Open access literature is free for everyone, not just for academics. There is no freeloader problem here.

OA through archives is so inexpensive for the host that there's no need to look for co-contributors to defray the costs. Moreover, institutions that host OA archives for their own research output benefit greatly from from the resulting visibility and impact.

OA through journals is more expensive. But once the journal's costs (peer review, manuscript preparation, dissemination) have been completely covered, it doesn't matter that some members of the audience have access without having paid. So there's no freeloader problem here either.

Think about TV and radio. They are free for users because the production costs were paid up front by someone else. Do TV and radio stations worry that some users are getting the broadcasts without having paid? No. The whole idea is to cover the costs from other sources so that users don't have to pay.

Another way is for authors to realize that they typically want two things from a journal publication: prestige and impact. Because most OA journals are young and haven't garnered prestige in proportion to their quality, authors seeking prestige generally seek it from conventional journals. But it's only a matter of time before the most prestigious OA journals are as prestigious as any others.

But for impact, no subscription journal can compare to OA journals. OA distribution gives authors a much wider audience than any subscription-based journal, and this larger audience translates into greater citation impact. This was well-documented by Steve Lawrence in this piece for Nature, http://www.nature.com/nature/debates/e-access/Articles/lawrence.html.

Authors can have both prestige and impact by depositing their articles in OA archives and publishing them in prestigious journals. So the solution is to remind authors of their full set of interests and show them how much is already within their power.

Then suppose you publish the same article in a journal that demands transfer of copyright. You can ask the journal to let you retain copyright; a growing number will comply. Or you can transfer copyright but ask for the right to deposit the postprint (the version approved by the peer-review process) in an OA archive. About 55% of journals (not journal publishers) already allow this, and a larger number will agree if asked. See the table of publisher policies on self-archiving and copyright maintained by Project Romeo, http://www.lboro.ac.uk/departments/ls/disresearch/romeo/. Some will let you put the postprint on your "personal web site" but not in an OA archive. There is so little difference from the user's standpoint that journals are silly to insist on this distinction; but if that's all you can get, take it. Google will pick it up and help others find it.

If the journal refuses, at least you still have the preprint online. If you want to, you can supplement the preprint with a list of the differences between the preprint and postprint, a document not covered by the copyright in the postprint that you transferred to the journal. But it's almost never necessary to go that far.

In my view the Ingelfinger Rule protects a journal's profits much more than it protects public health. There is *far* more danger to public health from the health-related rumor-mongering and superstition already rampant on the internet than from research articles by professionals awaiting review. Moreover, many of these researchers have already survived one layer of peer review, to get the grant that funded their research. Finally, of course, peer review does not eliminate error either.

A competent OA archive displays the metadata about its articles, letting users know when an article has been peer-reviewed and when not. This further reduces the risk to the public. Moreover, any real remaining risk must be balanced against the huge gain of widening access and shortening delays to this very important body of research. For example, the genome of the SARS virus was sequenced in about seven days thanks to the open sharing of data and results.

A growing number of journals have stopped using the Ingelfinger Rule. What we really need is good information about which journals still use it and which don't, so that researchers can make better judgments about whether they can self-archive their preprints. Besides, once we know where the rule is still in force, we know where to send letters of protest and complaint.

We don't know how this will play out. One plausible view is that we'll see a market in processing fees. Authors will start to consider the size of the fees in addition to the prestige, quality, and usage of a journal, in deciding where to submit their work. If so, that will be a constraint keeping fees down. Or, if a journal keeps its fee high, it's only because it can still attract submissions. Since the scholarly journal industry is notorious for the absence of competetition for subscribers (one journal doesn't publish another's papers, even if its price is lower), this would introduce the first meaningful competition into the journal marketplace. Or, it would replace a dysfunctional competition for subscribers with a real competition for authors.

Another plausible theory is that journals with very high rejection rates are very prestigious and likely to survive intact in a new market otherwise dominated by OA journals. So these journals would never convert to OA and never charge the very high fees.

I think both are likely to happen. We'll see OA processing fees fluctuate according to something much like market forces. We'll also see priced and OA journals coexist.

But I am curious. What would an author ever want to do with a research article that a traditional journal would permit and an OA journal not permit? The reality is the other way around. Once you have transferred copyright to a journal, you have lost ownership, lost control, and lost rights. Often you can't post it to your own web site or make copies for colleagues without asking permission of the new owner. But with an OA journal you can do all of these things and many more.

One key point: consenting to OA for your work is *not* the same as putting it in the public domain. You waive some of your rights and retain others. You can retain any that you like, of course, and any that the journal will let you retain. At most OA journals, for example, you retain the right to block the distribution of copies that mangle or misattribute your work. Some OA licenses also give you the right to block commercial reuse. But with OA, you consent to the uses that promote scholarship, inquiry and research (reading, downloading, copying, sharing, storing, printing, searching, linking, and crawling) and retain the rest. With conventional journals, you transfer the whole bundle of rights and retain nothing.

If you have further questions about open access, don't forget to check out the article that accompanied this Colloquy Live, any of the links that Peter provided at the beginning of the conversation, or his blog and newsletter. Thanks for joining us today!