Uncovering open access | Opensource.com

Uncovering open access

Posted 09 Nov 2010 by 

Rating: 
(5 votes)
Image by : 

opensource.com

submit to reddit

To the general public, “doing science” is all about discovery. But in truth, that’s only half the picture. Consider the experience of an obscure nineteenth-century Augustinian monk...

From 1856 to 1863, Gregor Mendel cultivated and observed 29,000 pea plants and managed to unlock some of the secrets of heredity, including the concepts of dominant and recessive traits.

In 1865, Mendel presented his findings as a two-part lecture, “Experiments on Plant Hybridization,” before the tiny Natural History Society of Brünn (present-day Brno, Czech Republic). A year later, he published his findings in the society’s Proceedings, of which 115 copies are known to have been distributed. With that, his painstaking work disappeared—virtually without a trace—for 35 years. In scientific terms, an eon.

Biologists struggled on, fruitlessly seeking to explain heredity through blending theories or with Darwin’s earnest but wrongheaded notion of “pangenesis.” At last, in 1900, Mendel’s work was rediscovered and helped spark the modern science of genetics. The fate of Mendel’s research is a sharp reminder that aside from discovery—i.e., the findings—science is also emphatically about dissemination, that is, access to (and application of) those findings.

Open access seeks, through the power of the Internet, to make scholarly materials freely available to the world. No passwords. No subscription fees. And today, it is perhaps the hottest flashpoint in science publication—with Harvard, and an unlikely firebrand named Stuart Shieber ’81, right in the thick of things.

Shieber, a Harvard School of Engineering and Applied Sciences (SEAS) computer scientist and tireless advocate of open access, crafted an open-access policy for scholarship that the Harvard faculty of arts and sciences (FAS) adopted, unanimously, in February 2008. thoughtful, soft-spoken, almost serene, he hardly seems suited to the role of revolutionary.

Richard Poynder, an astute observer of the changing landscape of journalism and publishing, suggested in his blog that this “poster child” for open access might lack the necessary grit to push the movement forward. Undeterred, Shieber keeps on pushing.

The heart of his proposal: “Each Faculty member grants to the President and Fellows of Harvard College permission to make available his or her scholarly articles and to exercise the copyright in those articles.” Don’t be fooled by the matter-of-fact language: this is pure dynamite. It grabbed headlines and roiled the Internet, hailed by some as “bold and visionary” even as a “shot heard ’round the academic world.”

While other universities’ open-access policies had been premised on individual faculty members opting in, in Harvard’s policy, the presumption is that faculty will make their work freely available, unless—for a given publication—a faculty member opts out.

According to Peter Suber, a leading proponent of open access and currently a visiting fellow at the Law School’s Berkman Center for Internet and Society, it’s “the best university policy anywhere.” A newly created Office for Scholarly Communications (OSC), headed by Shieber, was tasked with administering the policy, which includes archiving of publications in a central repository, known as DASH, or Digital Access to Scholarship at Harvard.

But why all the fuss? Hasn’t the Internet already put worlds of knowledge just a mouse-click away? Shieber points out that, contrary to conventional wisdom, academics’ access to relevant scholarly information is actually decreasing. The problem isn’t with the technology for disseminating knowledge but rather with the increasingly inability of institutions and individuals to pay for access. Harvard’s policy seems almost elegant in its simplicity. “Under the new policy,” Suber explains, “faculty members retain some of the rights they formerly gave to publishers, and they use those rights to authorize open access.”

For academic publishers, it was an unprecedented shot across the bow. Moreover, the School of Education and the Law School quickly adopted similar policies. And on September 15, Cornell University, Dartmouth College, MIT, and the University of California at Berkeley all joined Harvard in a general compact to support open-access publishing by providing administrative, technical, and monetary support.

Change is in the air, but where is it headed—and what does it portend? To understand the controversy over open access you need to untangle a complex back-story, one that involves the long history of the academy, the rise of scholarly publishing, and the federal funding of research since the Second World War—with side trips into the rise of the Internet and the changing role of libraries. Even human (or, at least, faculty) nature figures into the mix. It’s complicated but essential to understanding where we are …and where we’re likely to be headed.

Movable type breaches the ivory tower

Prior to the Enlightenment, the university was a very different place from today’s research-centered institution. Medieval universities, such as Oxford (ca. 1167), had a focus that was largely religious, and for centuries their efforts at dissemination likewise stressed religion: Harvard, for example, was founded in 1636 mainly for the purpose of training ministers. But gradually, under the influence of humanist principles that emerged during the Renaissance, the university began a metamorphosis that would ultimately yield today’s modern research institution. And publishing lay at the heart of the transformation.

Oxford University Press (OUP) was one of the very earliest modern academic publishers. As the OUP website explains, it emerged as part of “the information technology revolution of the late fifteenth century, which began with the invention of printing from movable type.” Lead letters, tucked into wooden shadowboxes. There’s something charming about linking the digital revolution back to a collection of smudged metal blocks. But compare the arduous task of hand-copying and illuminating a manuscript with the relative ease of setting (and editing) type, and suddenly it makes sense as a game-changer.

Moreover, with the widespread adoption of modern printing, the gates of academia began to crack open. Universities began by publishing Bibles and other religious works but soon the scope of their activity began to expand, diversifying into dictionaries, biographies, music, and journals. Universities learned that to preserve the knowledge of their faculty, they needed to disseminate it: knowledge and access to knowledge went hand in hand. Thus universities also turned to publishing academic materials, and by the eighteenth century, many were publishing their own research journals. 

Despite the growth of university presses and the burgeoning of scientific activity, research findings continued to be surprisingly inaccessible.

Commercial publishers tiptoe onto the scene

In contrast, commercial publishers entered the scholarly landscape more gradually. Then as now, commercial publishing was highly competitive, and publishers, highly risk-averse. In a perceptive essay on American university presses, Peter Givler observed that leaving the publication of “scholarly, highly specialized research” to the commercial marketplace “would be, in effect, to condemn it to languish unseen.” Commercial publishers, however, had the necessary distributional muscle.

Despite the growth of university presses and the burgeoning of scientific activity, research findings continued to be surprisingly inaccessible. Well into the nineteenth century, much of the exchange of scientific information took place behind closed doors—in Victorian lounges of members-only organizations, like the Royal Society in England, or in the growing number of local natural history or scientific societies. Though many of the local societies published Proceedings, they and their publications actually limited access to research findings—as with Mendel and the 115 known copies of the 1866 Proceedings of the Natural History Society of Brünn.

Such efforts at least reflected an awareness of the need to disseminate knowledge. In the United States, for example, the American Association for the Advancement of Science (AAAS) was founded in 1848. And the Lawrence Scientific School at Harvard (the progenitor of SEAS) was established in 1846−47. In a letter to the University, Abbott Lawrence—whose gift funded the new school—voiced his concern “that we have been somewhat neglectful in the cultivation and encouragement of the scientific portion of our national economy.” Among those neglecting the nation’s scientific enterprise were, with few exceptions, publishers. But by the late nineteenth century, commercial scholarly publishing began to emerge in something approaching its modern sense. In 1869, for example, publisher Alexander Macmillan founded Nature in Britain. (For some time, it survived mainly as a personal labor of love: according to the Nature website, Macmillan “tolerated a loss-making venture for three decades.”)

In the United States, Science Magazine (later simply Science), played a similar role as a commercial journal for broad-based scientific knowledge. And, like Nature, it faced economic challenges. Backed by Thomas Edison and Alexander Graham Bell, the journal began rolling off the presses in 1880. But it barely survived a series of financial crises before gaining a measure of stability by partnering with the nascent AAAS. Science and Nature elevated, even liberated, scholarly communication, presenting a broad overview of discoveries to anyone willing to pay the then relatively small subscription fee. Moreover, many early commercial scholarly publishers, like Macmillan, valued the dissemination of knowledge as well as profit-making.

But what’s the relevance of Victorian publishing efforts to open-access scholarship? The issues that were present at the outset of modern scholarship and scholarly publishing—how to distribute and archive knowledge and, no less important, how to pay for it—haven’t changed. They’re the same challenges that we confront today. From hand-illuminated, gilt-edged manuscripts to lead type and steam-powered presses to silicon and bits and bytes, the issue has always been access—who has it, what does it cost, and is it sufficient?

Scholars and publishers—a partnership in peril

By the latter half of the twentieth century—as scientific enterprise and college and university enrollments burgeoned—it appeared that researchers and publishers had achieved a lasting partnership. Much of the credit, ironically, goes to the Second World War. The war highlighted the strategic importance not just of science but of access to scientific information. Early on, Pres. Franklin D. Roosevelt and Prime Minister Winston Churchill reached an informal agreement that the United States and Great Britain would share (at no cost) any scientific development of potential military value. The federal government also collaborated as never before with universities and the private sector, above all in the Manhattan Project and the building of the atomic bomb but also in a host of other developments including radar, sonar, synthetic rubber, nylon, the proximity fuse, and napalm (developed in 1942−43 by a Harvard team headed by chemistry professor Louis F. Fieser). This federal engagement in scientific research didn’t end with peacetime; indeed, it continued to expand.

Thus, Edwin Purcell, Gerhard Gade University Professor, Emeritus, and co-winner of the 1952 Nobel Prize in Physics, was instrumental in developing the principles of nuclear magnetic resonance (NMR). In the coming decades, the applied sciences continued to blossom. At Harvard alone, Harold Thomas, Jr., whose last position was as Gordon McKay Professor of Civil and Sanitary Engineering, spearheaded the famed Harvard Water Program; Ivan Sutherland conducted research that resulted in the first head-mounted display, one of the first attempts at virtual reality; and the University became an early node on the ARPANET, the precursor of the Internet.

Taking advantage of the bustling research activity, economies of scale offered by advances in printing technology, and their well-established editorial and production skills, commercial publishers now seized on science as a viable, paying venture. New fields blossomed—computer science, cognitive science, neuroscience—each accompanied by the requisite specialty journals. In many ways, publishers (and particularly the editors of such journals) joined academics as partners in the scholarly enterprise. Faculty members provided the content; publishers lined up volunteer peer-reviewers, arranged for promotion and distribution, and helped hone and polish the manuscripts. Indeed, scholarly publishers in many respects helped to shape fields. They generally also demanded that authors cede their entire copyright interest. (Publishers, however, still pay nothing for these rights.)

The partnership was seen as analogous to the role of art museums. If people wanted to see the paintings, they had to pay to enter. And if artists wanted others to see their art, they needed to entrust that work to the gallery or museum. Given the small number of subscribers and the high quality of the value-added editorial work, the cost of access seemed justified. And before the rise of instant online publishing, how else could academics circulate their work in a form more durable than an oral presentation? Particularly in the last 20 years, Shieber says, “demand has been inelastic, and [the publishers] clearly took strong advantage of that.” Which was fine as far as academics were concerned: they “never really knew the direct costs” because university libraries picked up the tab.

Unfortunately, for both academics and publishers, the honeymoon was short-lived. Well before the current economic storm, the pricing model for scholarly publication was beginning to break down. Publishers charged ever-higher subscription fees. (Institutional online subscriptions to journals like Brain Research can now run as high as $20,000 a year.) Libraries and universities bridled at the rising costs but initially did little to push back. While individual institutions might negotiate better or modified deals with publishers, overall, libraries have found it a losing game.

For example, in 2007 the Max Planck Institute stopped subscribing to Springer’s journals in protest of the high cost, but, as one version of the story goes, researchers demanded the return of their journal collections. A year later the Institute resubscribed after negotiating a time-limited “experimental” mix of open access and subscription models with Springer. Under the agreement, all authors from Max Planck gained access to 1200 journals and had the costs waived for Springer’s Open Choice program that “offers authors to have their journal articles made available with full open access in exchange for payment of a basic fee (‘article processing charge’).” A sign of progress for sure, but a limited one as Max Planck still had to foot a substantial bill (the financial terms were not disclosed) and other institutions and readers remain on the hook for substantial subscription fees, as the vast majority of Springer’s content remains locked.

In fact, prohibitive subscription costs and, more recently, budget cuts have forced many libraries to permanently cut back on journal subscriptions, online and print. Even in the best of times, Harvard (which maintains one of the largest libraries in the world) hasn’t been able to subscribe to every journal. Today, however, the situation is significantly worse. One problem, explains Martha “Marce” Wooster, head of SEAS’s Gordon McKay Library, is the lack of any “algorithm a librarian can use to determine what journals to keep or cut,” whether based on price or need for access. In some cases, publishers now tie multiple journals together, taking advantage of online portals. And while you can often hand-tailor journal subscriptions for your institution (only requesting access to the particular journals faculty would need), the overall cost savings relative to buying the entire package is limited or nonexistent. The upshot is that—aside from interlibrary loan or direct correspondence with researchers—the only way for a faculty member to see certain research findings is by paying the publisher’s going rate. As libraries face continued cutbacks, this will present more and more of a challenge for the scholarly enterprise.

Rethinking the model

The Internet, of course, was the final kink in the chain of events that imperiled the academic−commercial publishing partnership. That might not seem surprising. The Web has radically revised much of life, including academia. At SEAS, for example, CS 50, “Introduction to Computer Science I,” and QR 48, “Bits,” are available online. And instead of a textbook, the introductory Life Sciences sequence now makes use of a multimedia showcase. MIT’s open courseware initiative is a permanent, ongoing program to make available—free and online—much of MIT’s course content. But curiously, while we can accomplish more and more online (sending holiday cards, managing our bank accounts, renewing driver’s licenses, etc.), the knowledge produced by research universities like Harvard has remained comparatively inaccessible, in a sort of cyberspace lockdown. How did that happen? How was it allowed to happen?

Before its emergence as a full-fledged social networking and shopping tool, the Internet served university researchers. But if the current commercial model for publication goes unchallenged (leaving publishers as gatekeepers to Web-based scholarship), the Internet could become more a barrier than a catalyst for the scholarly enterprise. “If the publisher owns and controls the scholarly content,” Shieber says, “there’s no way to prevent that publisher from limiting access and charging for that access.” In the early days of the Internet, the cost of digitizing print materials was far from trivial, and publishers could justify high online-subscription fees as they migrated journal content to this uncertain new medium. Thus, while many other sites began giving away content, scholarly publishers retained and further beefed up their access controls.

But are there viable alternatives? Any solution will require a rethinking of the status quo in scholarly publishing. One possibility is open-access journals, exemplified by the publications of PLoS (Public Library of Science), founded in 2000 by Nobel laureate Harold Varmus. In PLoS journals, authors retain all rights to their work, and anyone can download and use information from PLoS free of charge, provided proper credit is given. But such efforts are still dwarfed by the subscription journals, and some of the traditional, longer-standing publications still “count for more” in terms of academic status. Moreover, open-access undertakings aren’t cost-free: such journals still need to be supported and maintained.

In the case of many open-access journals, these costs are shifted to the author’s rather than the readers’ side. On the highest end, a flagship PLoS journal, for example, charges authors around $3,000 for the publication of an article. That makes subscription-based publications, which typically don’t charge authors for publication, much more attractive for many authors.

In addition, it’s important to appreciate the value added by commercial publishers. Particularly in an age of instant publication, Shieber argues, “you need people that do what editors and journals do now.” This includes managing the peer-review process, performing editorial and production work, and distributing and archiving the final product. No less important, “there’s the branding and imprimatur,” something that happens “more or less as a side effect” of publication in a particular journal.

The Web has made it possible to “unbundle” these activities: peer review could happen in an online forum; editorial and production work could be done almost anywhere. But the activities themselves (and their coordination) remain essential, and they come at a price. Unfortunately, Wooster says, Web technology has yet to put scholarly publishing “on an economic basis that’s not dysfunctional.” As counterintuitive as it sounds, the shift of scholarship from print to Web (unlike music’s move from CDs to MP3s) has resulted in significantly higher prices, despite doing away with a raft of standard printing costs. Shieber felt there simply had to be another way, where scholarship ends up winning.

The man with the plan

With powerful incentives for publishing in commercial journals, why would an aspiring faculty member choose open access? This is where Shieber steps in. His plan is to make open access viable, in essence, by leveling the playing field for scholarly publication. According to Shieber, all that’s needed to set open access on par with commercial options is that “those underwriting the publisher’s services for subscription-fee journals commit to a simple ‘compact’ guaranteeing their willingness to underwrite them for processing-fee journals as well.”

This commitment, if widely embraced by universities and funding agencies, would reduce the disincentives for faculty to publish in open-access venues. While supporting open-access publishing fees might seem like a costly commitment for an institution to take on, in some scenarios, it could be completely offset by the reduction over time in fees for subscription-based journals. Indeed, existing subscription journals might even convert to open-access models (should the benefits be sufficiently compelling). Shieber’s compact has another advantage: paying to publish is more transparent. Traditionally, Shieber writes, “consumers of scholarly articles have been well-insulated from the cost of reading.” If faculty were brought face to face with the actual costs, they might more easily be enlisted in an effort to address the problem.

To Shieber, such a compact would be a rational, fair, and economically sustainable way to support the publication of scholarship. Others aren’t so sure. One natural reaction would be: “Why should there be any problem—or compact—at all? In this age, why should anyone have to pay to have access to scholarship?” Stevan Harnad, an outspoken “archivangelist” for giving scholars the green light to freely and immediately archive any article they author and a professor of cognitive science at the University of Southampton (UK) School of Electronics and Computer Science, sums up this argument in a haiku:

It’s the online age
You’re losing research impact...
Make it free online.

Harnad believes that anything that distracts from his goal of achieving universal access in short order (opt-in policies; definitions of copyright; or simply dealing with the various rules or fees publishers request for granting open access) distracts us from the real issues. And if it does, Harnad contends, the research community “will yet again have shot itself in the foot insofar as universal OA [open access] ...is concerned” Rather than rearranging the deck chairs on the Titanic, he says, why not do something truly transformative. Harnad’s paradigm: open-access self-archiving, in which authors deposit material in open digital archives. In short, even right-minded publishers and supportive academic institutions often get in the way of scholars reaching their audiences directly. 

Shieber contends that unless the incentives are right, allowing authors and publishers to choose among a variety of publishing platforms without penalty, scholarship will continue to suffer until a crisis is reached.

Shieber answers such criticisms much as an engineer would. Many aspects of commercial publishing work well, he says. Instead of battling the system, our aim should be to set it on a realistic foundation. “You can be passionate about all kinds of things,” he says, “but if there’s no way for the economics to work out, it won’t work.” Notions that “we should just give everything away” or that “information wants to be free” aren’t real-world solutions. The free-content model, even with the help of advertisements, probably isn’t sustainable in the short term, let alone the long. (The newspaper industry can certainly corroborate this.) While praising the good intentions of open-access advocates like Harnad, Shieber warns that the situation is far from simple. “There’s a notion that once we’ve solved the access issue, our problems are over,” Shieber says. The reality, he fears, won’t be so obliging.

A case study from the American Physical Society (APS) highlights this point. In an article in the November 1996 APS newsletter, Paul Ginsparg (now a professor of physics at Cornell) observed that:

publishers had defined themselves in terms of production and distribution, roles which we now regard as largely automated … [T]he essential question at this point is no longer whether the scientific research literature will migrate to fully electronic dissemination, but rather how quickly this transition will take place now that all of the requisite tools are on-line.

Ginsparg suggested that a shift to electronic dissemination would quickly resolve the access problem. But more than a decade later, with scientific journals duly ensconced in cyberspace, the issue of access continues to be thorny and unresolved. “People are economic actors,” Shieber says, “and that means they work in their own interest, subject to whatever constraints they’re under.” For the publishers, owning the copyright to published articles and restricting access (through high fees) improves the chances of making scholarly publications pay off. And what of the academics? While it’s in scholars’ best interests to have the broadest possible access to their work, the rewards (and familiarity) of the existing system exert a strong gravitational pull.

Shieber contends that unless the incentives are right, allowing authors and publishers to choose among a variety of publishing platforms without penalty, scholarship will continue to suffer until a crisis is reached. “While you can’t separate the economics from the access issue,” he says, “the economic issue is clearly secondary.” Once the economics are taken care of, faculty will be able to focus on taking scholarship to the next level. And it’s there, in discussing the university’s role as a gateway to knowledge, that he truly lets his passion show. “The whole point of the university is that we’re supposed to be engaged in the generation of knowledge for the good of society,” he says. “So shouldn’t society be able to get the goods?”

Scholarship as a public good... and public goods

Shieber’s focus on economics and practicalities undoubtedly has merit and fits the SEAS emphasis on “what works” and “practical applications.” But there are other factors involved here: open access raises matters of principle, both philosophical and political. Absent open access, how effectively can scholarship promote the public good? To some—i.e., those who find it hard to imagine “scholarship” and “the public good” in the same sentence—limited access to an article on the Casimir effect or yet another interpretation of James Joyce’s Ulysses wouldn’t exactly qualify as a hot-button issue. Strictly ivory tower stuff.

In the case of science, engineering, and medicine, people generally understand that basic, seemingly esoteric research can yield profound improvements in our lives: NMR gives us magnetic resonance imaging (MRI) and improved medical diagnoses; digital and laser technologies give us CDs and DVDs; and a rotting cantaloupe in a Peoria, Ill., research lab gives us penicillin. Wooster wonders if future benefits may be threatened by the ongoing constriction of scholarly information. “If no one knows about a finding,” she says, “that’s a great disservice to the whole world. If universities are really aiming to improve the world, we need to think hard about this.”

And there are other matters of principle involved. As Wooster says, “It’s always been insane that the faculty do the research and write the papers and then give it all away—and we’re forced to buy it back. … I mean, don’t we already own it?” Others make a similar argument with respect to taxpayers: they’re being denied something they’ve already paid for. After all, federal funds support a significant portion of university research. (In 2009, federal funding underwrote about 80% of all SEAS research.) But, with few exceptions, the rights to this publicly funded research—when published in a traditional academic journal—are transferred gratis to the publisher. Regardless whether taxpayers are interested in, say, reading up on the latest advances in fuel-cell technology, Shieber and other open-access advocates insist they should be able to do so: the principle still applies.

Such considerations suggest another valid model for open-access scholarship, aside from self-archiving and Shieber’s compact. The federal government could step into the breach: after all, what’s at stake isn’t merely the “public good” but public goods. Given that much research is publicly funded, the people (and their government) have an interest in the results (and access to those results). In other words, the researcher-publisher partnership isn’t bipartite—it’s a “three-way street.” Perhaps all federal research grants could stipulate that findings achieved with public support must be made freely available, either through publication in open-access journals or, if published in subscription-based ones, by being made concurrently available via a free digital archive. At the same time, federal research grants could cover reasonable publication costs. The result, with some tweaking, could reconcile the interests of all parties.

In fact, we already have a working model for a good part of this approach, in the area of medical research. The National Library of Medicine, part of the National Institutes of Health (NIH), operates PubMed, a comprehensive database of journal citations and abstracts. It also maintains PubMed Central, a “free digital archive of biomedical and life sciences journal literature.” In April 2008, at the instruction of Congress, the NIH adopted a new public-access policy, mandating that any investigator funded by the NIH must submit a copy of any peer-reviewed manuscript that had been accepted for publication for posting in PubMed Central. At present, the policy doesn’t provide funds to cover costs associated with publication, somewhat limiting this best-practice scenario.

Other ways to open access: Brave new libraries and digital archives

The devil, of course, is in the details, and questions remain about the likely fallout from any programmatic commitment to open access. Would academic presses and parts of the commercial publishing business perish? Shieber insists that his goal (and the overall intent of open access) has never been “to take down the publishers.” Especially in the Wild West environment of cyberspace (from blog postings to tweets to suspect Wikipedia entries), editors and peer review are as essential as ever.

For his part, Shieber downplays the impact of the current debate. “A very small percentage of written works would fall under the open-access regime as it’s currently being discussed,” he says. “Right now, we’re just talking about scholarly articles—peer-reviewed, specialist works.” The discussion doesn’t involve books (see online content), editorials, content written by journalists, etc. Even if scholarly research shifted entirely to open-access journals tomorrow, publishers would hardly be out of work. In any case, there will probably be an enduring market for print versions of publications. (Kindle has yet to displace Amazon’s book department.)

But what about the impact on libraries? As more and more collections move online, will the library’s storage function be diminished? Many people assume that moving to e-journals and digital collections would pose a threat to libraries, with the latter winding up nearly invisible. Wooster strongly demurs. She doesn’t think the digital library will be any less vital than today’s paper-chase repositories. John Palfrey—the Henry N. Ess III Professor of Law, head of the Law School library, and co-director of the Berkman Center for Internet and Society—anticipates the emergence of a whole new type of librarian (his preferred term is “empiricist”) as libraries transform into information centers, with open-access archives playing a major role. Much like publishers, libraries offer services that will continue to be of value even if journals go online and open-access archives become commonplace. Shieber agrees. “Library services (reference, teaching, and new functions designed to make open-access materials available) will all continue to be needed and will be incorporated into the library’s purview,” he says. And, Wooster points out, the library’s role as a place for “a record of history” isn’t likely to be changing any time soon.

Moving forward on the question of access may also mean looking back to the traditional role of university presses as editors and distributors of a given institution’s scholarship. (Shieber concedes the irony that he’s published his books with MIT Press rather than Harvard University Press.)

University-based digital archives represent another avenue for open access. Such archives are already coming into being, as seen in the DASH open-access repository or the Medical School’s Harvard Catalyst project, which features a social networking database for linking researchers and research (including links to the DASH archive) across the entire University. And some fields (physics, for example) have long maintained free archives of preprints, for example, arXiv.org, developed by Cornell’s Ginsparg.

Mobilizing Harvard faculty behind the open-access model and developing a practical platform for its implementation will be ongoing challenges. Shieber has invited fellow open-access expert and Berkman Center Visiting Fellow Suber to help further foster the process. Shieber cautions that any shift to open access will take time. Even after the conceptual and economic issues are resolved, he doesn’t advocate changing everything at once. But once the issue of open access becomes secondary and the claims of researchers and publishers have been reconciled in a fair (and sustainable) way, then, he believes, scholars, universities, publishers, and the wider world will be able to focus on a bracing new realm of discovery. And isn’t that what we really want science and research to be about?

Further reading

Learn more about Harvard's work with open access:

This article originally appeared in the Harvard SEAS winter 2010 newsletter and was reprinted with permission from the authors, who have agreed to release it to the public under a Creative Commons (CC BY SA) license.

submit to reddit

3 Comments

G. Hexner

I have written several scientific papers, and I would like to present my perspective on "Open Access". The article ignores a key point in the open vs. traditional publishing. The readership and the impact f an article is very much influenced by where it is published. All of the well prestigious journals are of the traditional type, protected by pay walls. Every author would like his work to appear in the most prestigious journal possible. He is then faced with publishing his work in an unknown, but open access; or in a well known, but closed journal. There is in fact a further difficulty. My research was carried out without the benefit of a research grant. I therefore do not have access to a fund, which covers page charges. My articles have so far appeared in the traditional journals, and I so far have never paid for their "voluntary page charges". Even in the traditional world, these charges are non-trivial, they can reach $1000.00 In the open world, according to the model suggested, a publishing charge would be compulsory, so it would be problem for me.
There is a further contradiction built into the traditional model. The editors of only some of the journals are salaried positions. In contrast, the reviewers are always volunteers, taken from the same pool of people, who write the articles in the journals. When they get an article to review, they need access to the references sited by the article. For those, like myself not affiliated with a major university this presents a problem, as the journal do not grant access even to their own archives for the reviewers!
I hope that the open access model succeeds, as it is in the interest of every author to reach the widest possible audience. However, the model needs to be amended to include people such as I, who do not have access to a fund that covers publication charges.

Vote up!
0
Vote down!
0
amsciforum
Newbie

SUMMARY: Universities need to commit to mandating Green OA self-archiving before committing to spend their scarce available funds to pay for Gold OA publishing. Most of the university's potential funds to pay Gold OA publishing fees are currently committed to paying their annual journal subscription fees, which are thereby covering the costs of publication already. Pre-emptively committing to pay Gold OA publication fees over and above paying subscription fees will only provide OA for a small fraction of a university's total research article output; Green OA mandates will provide OA for all of it.

Journal subscriptions cannot be cancelled unless the journals' contents are otherwise accessible to a university's users. (In addition, the very same scarcity of funds that makes pre-emptive Gold OA payment for journal articles today premature and ineffectual also makes Gold OA payment for monographs unaffordable, because the university funds already committed to journal subscriptions today are making even the purchase of a single print copy of incoming monographs for the library prohibitive, let alone making Gold OA publication fees for outgoing monographs affordable.)

Universal Green OA mandates will make the final peer-reviewed drafts of all journal articles freely accessible to all would-be users online, thereby not only providing universal OA, but opening the doors to an eventual transition to universal Gold OA if and when universities then go on to cancel subscriptions, releasing those committed funds to pay the publishing costs of Gold OA.

Vote up!
0
Vote down!
0
Brice

As you know, it is not always easy to find scientific articles in free full text.
For this reason, with google technology, i have developed a custom search engine in order to find free full text scientific articles in PDF Format. This search engine indexes more than 10 millions of free references (mostly to journal articles, conference papers and technical reports).

You can try it at this address: http://goo.gl/248em

Please contact me if you have free sources of scientific articles (Open archive …)

Vote up!
0
Vote down!
0

Michael Patrick Rutter is Communications Director at the Harvard School of Engineering and Applied Sciences (SEAS). Before joining SEAS, Michael spent a decade working in academic publishing.

James Clyde Sellman is a freelance writer and editor whose work has appeared in Harvard Women’s Health Watch, Harvard Health Letter, and About.com’s Health Channel, among others.

What is open education?

Holiday gift guide promo