In 1665, Henry Oldenburg, a German publisher and prolific letter writer, changed the future of science. His Philosophical Transactions of the Royal Society of London spawned the dominant form of scientific communication for the next three and half centuries. Robert Hooke and Isaac Newton waged intellectual battles over springs, optics, and gravity in its pages. By the late nineteenth century, other journals, including Nature and Physical Review, appeared, and as scientific progress marched along, their pages filled with articles by Albert Einstein, Marie Curie and other scientific luminaries.

Journals hold a special place in the lives of scientists. The number of articles they publish and in what journals significantly affects the course of their careers – where they work, how much funding they get, and how much respect they can expect from their peers. For some, articles are not just a means of communication, but end products. Today, however, business pressures and emerging technologies seem likely to change everything about scientific communication – who finances it, who has access to it, even what constitutes a scientific journal.

These journals have operated under the same basic rules since their inception. Authors submit research for publication to the most prestigious journal. Their colleagues, commonly three, read and review the article anonymously, and recommend to the editor to either accept or reject the article for publication, as well as noting its strengths and weaknesses, and suggesting revisions. This peer-review does not certify that these decisions are good ones – the reviewers do not replicate the experiments or analyze the data. It only means that the methodology is sound, the conclusions follow the data, and the research has some degree of relevance. Once an article passes peer review, the journal publishes it months or even years later, and other scientists in that particular field eventually read it. In recent times, some journal editors have taken steps to accelerate publication, such as posting articles online before they appear in print.

Financial problems, however, have recently plagued the scientific publishing industry. Non-profit and academic organizations used to dominate the business, but for-profit publishers entered the industry during the research boom of World War II. Today they publish two million articles a year with $8 billion in annual revenue. Most cost hundreds to thousands of dollars for an annual subscription, but some, like Brain Research and Combustion Science and Technology cost nearly $20,000 per year. This forces all but the wealthiest universities and foundations to cull their subscription list. Journals are expensive to produce, no matter how many subscribers they have. As the number of subscribers goes down, prices rise for the fewer institutions that must share the total cost of publication. In a vicious cycle, rising prices induce even more subscribers to cancel. This cycle applies to all types of journals, although the more specialized the audience, the more significant the impact. If one hundred libraries cancel their subscription to Nature, it barely affects the journal’s bottom line, but it might be the death knell for Colloid and Surface Science.

Shrinking access to scientific journals angers researchers the world over, as the ever-increasing costs of print publications shut out scientists from poorer institutions. Researchers’ frustrations climaxed in 2001, when 30,000 scientists from 180 countries signed a boycott letter calling on publishers to make their work freely available in online libraries within six months. They pledged not to “publish in, edit or review for, or personally subscribe to" journals that do not comply. “The current business model is unsustainable, since it is now economically and physically possible to disseminate an article worldwide at zero cost,” says Peter Suber, publisher of the Open Access Newsletter from the Scholarly Publishing and Academic Resources Coalition, an organization of universities and libraries that aims to raise awareness about this issue. “It’s a dysfunctional system; [publishers] are racing into a brick wall.”

Some changes have already taken place to satisfy researchers and keep traditional journals financially viable. Almost all journals published today put at least some of their articles on the internet, an “added value” meant to retain subscriptions by making the journals more worth their high cost in the eyes of buyers. Large databases link electronic articles, and scientists use search engines like “Web of Science” to gather, in a matter of minutes, the same information it previously would have taken months to find. Electronic versions of articles have quickly become indispensable to scientists. “I have journals going back for years on the shelves in my office,” says Vivian Siegal, executive director of a new open-access biology journal. “[But] if I need an article, I’ll get it online. It doesn’t even occur to me to read hard copies any more.”

However, just because an article is online doesn’t necessarily mean all researchers who want to read it can. Most scientists are affiliated with universities or research centers and collectively subscribe to journals through their institution’s libraries. While online subscriptions are sometimes cheaper than the expensive print format, publishers often bundle the two together, offering online subscriptions “free” to those libraries that purchase the print journals. In those circumstances, if scientists’ research institutions can’t afford the print format, then they can’t access electronic articles. This hurdle is prompting scientists to search for alternative ways to develop an open-access system, in which articles are available free, online, everywhere.

In recent years, open-access journals have made major gains in the publishing industry. The for-profit BioMed Central, a general collection of over 100 open-access journals, published its first article in 2000, with a novel pay-to-publish business model. BioMed Central makes money by charging a $1000 to $1500 processing fee to authors or their institutions and advertising aimed at researchers on its website. Also in 2000, the then-director of the National Institutes of Health, Harold Varmus, launched PubMed Central, a database in which publishers can deposit their articles six months to a year after original publication. Researchers can then access the database at no charge, while allowing traditional publishers to continue to profit. Over half of articles in medicine and health are freely available from PubMed Central, although not until well after they were initially published, which is still unsatisfactory to many researchers. In October 2003, the Public Library of Science, a non-profit organization of scientists and physicians advocating open-access scientific communication, launched its open access journal, PLoS Biology. One year later, PLoS Medicine’s first issue was published. Like BioMed Central, the PLoS journals charge a $1,500 processing fee to authors. As of October 2004, there are over 1,200 open-access peer-reviewed titles, more then twice as many as were available the year before. “It’s growing every day,” says Melissa Hagemann, program officer for the Open Society Institute, which funds open-access projects.

Open-access journals benefit almost all parties involved in scientific publishing. Authors gain increased visibility and prestige when more people can use the internet to read their work. And BioMed Central and the PLoS journals have committed to waiving the publication fee for those authors who can’t afford it. Readers, primarily other researchers, benefit when they have access to more articles. Theoretically, everyone wins when scientists can communicate more efficiently – research is done at a faster pace, discoveries are accelerated, and mistakes are uncovered more quickly. Since the public finances the bulk of American research through taxes, the argument goes, shouldn’t they too have a right to see what they’re paying for? Nobody wins under the current system except the “extraordinarily greedy publishers...who charge up to 10 times as much as the non-profits for an equivalent product,” says Paul Ginsparg, professor of physics at Cornell University and developer of the first large-scale database of scientific articles.

The path from the current journal system to an open-access one, however, has not been smooth. Some researchers fear that open-access journals will not implement quality control, i.e. peer review, in the same way as traditional print journals. A journal using the pay-to-publish business model might be more inclined to accept papers of questionable quality to keep up their revenue. Researchers also worry that funding scandals might be much more likely to occur when authors finance the publication of their work. “One of the fears is that open-access journals are going to somehow dilute the seriousness of our research,” says David Wood, an assistant professor at Princeton University studying bioengineering. “With free online journals, people perceive that you get what you pay for.” Open-access journals will need to reconcile the peer-review process with their business models before authors, universities, and funding agencies recognize them as acceptable venues for publication and agree to bear the burden of their cost.

Most publishers seem to agree that these changes are indeed likely to occur within the next 10 to15 years. Siegal predicts that 95 percent of all articles will be available freely through searchable databases within a decade, and that open-access journals will eventually replace their traditional print counterparts. While open-access journals will publish the bulk of day-to-day scientific research, she thinks there will always be a place for “research magazines” like Science and Nature. These journals, with a wide breadth of topics and a diverse paying audience, are taking a watchful waiting stance towards open-access publishing. “Switching to open-access now would be the demise of the journal,” says Katrina Kelner, deputy editor of Science. “We would have to do it cautiously and in a responsible way.”

Today, PLoS and BioMed Central look like online versions of traditional journals, but in the future, the whole concept of a journal may change. Scientists could simply deposit their papers in their institutional archival database. As long as they used compatible software to publish it, anyone with an internet connection could use a Google-like search engine to find it. A “journal” might be like a virtual Good Housekeeping Seal of Approval – it would tell readers something about the topic, how it relates to the field, and perhaps most importantly, its general quality, but would not physically exist as a bound collection of research papers. Also like PLoS and BioMed Central journals, the next wave of open-access journals will likely implement peer review in the traditional way. Eventually, publishers and scientists will have to figure out how to use changing technologies to adapt the peer review tradition to journals of the future. In the next decade or two, utilitarian search engines and high-speed connections may replace the traditional glossy pages of the venerable scientific journal.

“I foresee a true knowledge network rather than simple ‘electronic publication’,” says Ginsparg. “Most of the technical pieces are already in place, but the sociological obstacles, as usual, are the most difficult to overcome.”