JSTOR's Journal-Archiving Service Makes Fans of Librarians and Scholars

But some worry about loss of paper copies and about cost issues

By SCOTT CARLSON

Four bookshelves, heavy with bound volumes, greet visitors at the door of JSTOR's office here. A quick scan

ALSO SEE: From Pages to Pixels

of the room shows black steel bookcases lining the walls, cramming the corners, narrowing the walkways. They're packed with tattered, sun-bleached, occasionally handsome copies of journals like Classical Philology, The Journal of Risk and Insurance, and College English.

All these will soon be cut up and given a new life as digital documents. JSTOR, a nonprofit, digital-archiving organization, has since 1994 collected full runs of venerable journals, scanned the pages into a database, and made the contents available online, at a price. The nonprofit company's mission is twofold: to preserve and maintain journal literature, and to make that material more accessible.

"It's fabulously successful," says Mark S. Sandler, a collection-development officer at the University of Michigan's library. The service, which archives articles from more than 160 journals, is available at 786 American institutions and at more than 250 libraries around the world, from Estonia to Malaysia to Pakistan.

Mr. Sandler says JSTOR has had Michigan's runs of The William and Mary Quarterly and The American Historical Review on long-term loans. The university's librarians feared at first that scholars would not be able to get immediate access to those paper copies, he says. "But the reality is, we've had very few requests for those titles [in print], and it has been reassuring that JSTOR has been a suitable substitute for most users most of the time."

JSTOR's nonprofit status has engendered trust and respect among both librarians and journal publishers, two groups often at odds. The service has brought about changes in the way research involving scholarly journals is done: Scholars and students can search entire runs of popular journals from their offices or homes.

But some vocal critics ask whether JSTOR's efforts have led libraries to prematurely purge printed back issues of journals in favor of untested, and perhaps impermanent, technology. And those who buy information services for consortia of college libraries complain that JSTOR's prices are too high and out of reach for smaller colleges.

At large institutions, at least, JSTOR is one of the most popular digital-archiving services available. It allows users to search for key words and terms through decades -- even centuries -- of scholarly material. JSTOR includes both a black-and-white image of every page and a transcript of the text; users see only the image, while the search engine uses the transcript.

It's the sort of technology that allows Fred R. Shapiro, a librarian at Yale University, to look into the history of particular terms and sayings in English. As editor of the forthcoming Yale Dictionary of Quotations, he has used JSTOR to find the earliest instances of such sayings as, "There's no such thing as a free lunch." Bartlett's Familiar Quotations attributes it to the economist Milton Friedman, from his 1975 book of the same title. But by plugging "free lunch" into JSTOR, Mr. Shapiro found the saying attributed to Alvin Hansen in a 1952 issue of the journal Ethics.

"The stuff that I'm doing with quotations is not a typical use of JSTOR, but it serves well to illustrate what you can do with the search capabilities," Mr. Shapiro says. The company's archiving specialists "have a tremendous attention to accuracy, so it really is a high-quality substitute to paper."

JSTOR was created for the kind of use it sees at Alice Lloyd College, an institution with limited resources in the Appalachian hills of Kentucky. There, Paul Yeary, an associate professor of chemistry, uses the digital archive to read back issues of Science and other journals in his field. The librarians, meanwhile, don't have to find precious shelf space for back issues of the journal. They routinely give them away or pitch them.

The digital archive has also satisfied publishers, who worried that subscriptions would fall once journals began becoming available electronically. JSTOR does not post current or recent issues of journals, so as not to undermine their subscription bases. And the company has agreed to keep confidential its data on the archival popularity of specific journals. Publishers feared that releasing such rankings would lead libraries to drop subscriptions to the least-popular titles.

Kevin M. Guthrie, president of JSTOR, says that at libraries' request, the company does not reveal users' habits to the publishers. Librarians don't want publishers to use that data to hone or accelerate their marketing efforts.

"We're big fans of JSTOR," says J. David Baldwin, managing editor at the Ecological Society of America. The group was one of the first publishers to sign on with JSTOR, and it has since arranged a deal that allows its members to search JSTOR for back issues of the society's journals. The group has been able to save money on its storage of back issues, he adds. It now keeps only about 100 copies of each issue (in a warehouse in Kansas), and recycles the rest.

Karen Hunter, senior vice president for strategy at the commercial publisher Elsevier Science, says that JSTOR offered to digitize six of Elsevier's 1,200 journals, but that the company decided to digitize the journals on its own. "We felt that having the back files on the same platform as the current material was most convenient for the user," she says, adding that the decision didn't have anything to do with a fear of losing profits. "I don't think that back files are a terribly profitable business for anybody."

JSTOR's successes please William G. Bowen, president of the Andrew W. Mellon Foundation. The notion of an electronic archive first occurred to him in 1993, while he sat on the Board of Trustees at Denison University. The campus library had run out of shelf space, and the president had asked the trustees to set aside money for another building. Mr. Bowen learned that back issues of scholarly journals ate up most of the space, and that the stacks were often disorganized and incomplete.

"I said, 'This is really crazy,'" he recalls. "Here we are, being asked to respond to a situation where the material is inaccessible and disorganized anyway. This must be a problem everywhere." Indeed it was, as he found after contacting other college libraries across the country.

The Mellon foundation began supporting a pilot program to digitize journals, and Mr. Bowen chose Mr. Guthrie to lead it. At the time, Mr. Guthrie was working at the foundation, studying the financial troubles of the New-York Historical Society. Aside from his work with nonprofit organizations, Mr. Guthrie had also been a high-tech entrepreneur and a former college-football star. He had helped to develop an electronic pad on which coaches could outline plays.

A few key decisions about JSTOR were made early on, Mr. Bowen and Mr. Guthrie say. The project would begin archiving groups of journals according to scholarly disciplines, not groups of publishers. The service's first market was for the most popular and important journals in history and Mr. Bowen's own discipline, economics. Second, JSTOR would use the then-emerging Internet to distribute material -- not CD-ROM's, as Mr. Bowen had originally conceived. And the service would offer the digitized journal pages as images rather than text, preserving the look and accuracy of the paper original.

Another early decision: The digital archive would be a nonprofit organization, a status that has become crucial in its negotiations with libraries and publishers alike. "Some people said to me, 'This is a really good idea -- you could make money on this and should set it up as a for-profit,'" Mr. Bowen says. "I was convinced that was a bad idea. We needed trust, among other things. We wanted people to understand that we were out to do good in the world, not make a profit."

With a $10-million annual budget, JSTOR pulls in enough -- about $5-million -- to support itself and the archiving of the journals it currently offers. The organization raises the rest from foundations and puts that money toward digitizing future collections of journals. The most recent additions to JSTOR's database include 23 journals covering ecology and botany, going as far back as an 1867 issue of American Naturalist. Mr. Guthrie says it cost roughly $1.3-million to digitize them.

JSTOR has tried to keep its rates just above what it needs to continue operating, he says. Its prices are set according to institution size: To use the Arts & Sciences I Collection, which includes 117 journals, major research universities pay $45,000 to start the service and $8,500 annually to maintain it. The smallest colleges pay $10,000 up front and $2,000 annually.

But that's still too much for small institutions, says Thomas J. Sanville, executive director of OhioLINK, a library consortium based in Columbus. It considered buying into JSTOR a few years ago, hoping to get a price discount through a large purchase for its member institutions. But small colleges in the consortium found that they wouldn't get as much out of the deal as larger ones.

"For the Ohio State branches, they want $2,000 a year," Mr. Sanville says. "On the other hand, for the University of Dayton, they only want $4,000 a year. Now, the relative value of JSTOR at the University of Dayton, with 10,000 students, and Ohio State at Marion, with 1,100 students, is not $4,000 to $2,000."

Small colleges should be the target of JSTOR's mission, he argues. "I don't see them delivering. Maybe they can't deliver what they said they were going to deliver at a price tag that we all find attractive."

Mr. Guthrie responds that library consortia were used to getting deals on bulk buys, but that he couldn't offer them deeper discounts on JSTOR's service without undermining its financial structure. (Alice Lloyd College and other small institutions in its region were able to afford the service through a grant from the Mellon foundation to the Appalachian College Association, covering the start-up fees and paying part of the annual fees for the first few years.)

Although JSTOR was created to help libraries create space in their paper archives, the prospect of storing or getting rid of back issues has been a prickly and difficult issue for many librarians, despite the existence of the digital archive. Some librarians at institutions in Michigan, for example, say they are reluctant to discard or store their old journals despite their access to JSTOR.

Linda Shirato, a librarian at Eastern Michigan University, which subscribes to JSTOR, says her library stores old journals in a vault within the building, where getting access to them is fairly easy. Although the prospect of storing journals far into the future is a "problem" and a "challenge," she says, discarding them is not an option. "We don't dare. The history department is upset enough that we put them into storage. Libraries have been getting a lot of flak for that lately."

"The periodicals library wants to get rid of some things that are in JSTOR, but we haven't done that yet," she adds. "When the time comes to get rid of them, there will be quite a battle." Faculty members, she notes, are still inclined to use paper copies of journals.

One of those ready to do battle is Nicholson Baker, the novelist who recently has attracted attention for his attacks on libraries' use of microfilm and digital media, including JSTOR. His recent book Double Fold decries the destruction of paper periodicals in favor of digital media and microfilm, and he cites JSTOR as an influence on this trend.

In an interview, Mr. Baker acknowledges that JSTOR is a great research tool -- "a wonderful supplement" to paper collections of journals.

"But it's being used to substitute for these things that libraries, in their wisdom, put on the shelves for all time," he says.

Mr. Baker warns that the use of JSTOR is leading to a mass trashing of journal copies. There might be hundreds of copies of a particular issue of, say, the American Economic Review at campus libraries, he argues -- but each library might decide that the given issue is so commonplace that it can safely be trashed. Pretty soon, most or all of the issues are gone. "That's what happened with newspapers and microfilm" -- everyone bought into microfilm and pitched the papers, figuring that the large libraries were saving them, he says. But they were not.

Mr. Baker doesn't have a lot of formal data to support his view. He points to one of JSTOR's own polls, reported on its World Wide Web site, which says 31 percent of subscribing libraries have already discarded journal titles or plan to discard them in the future (http://www.jstor.org/about/bvs.html).

But the lack of data is part of the problem, he says, pointing out that no one is coordinating or keeping track of these journal cleanouts.

He is concerned that more libraries will follow the route taken by Denison University, Mr. Bowen's alma mater, where JSTOR was conceived. The university's library has formed a partnership with four other institutions, all JSTOR subscribers, to purchase journal-storage space in a former public library.

The colleges compile the best and most complete set of a particular journal among the five of them and store it for sharing. They can then discard the other copies. Denison keeps an additional run of frequently used paper journals, but it discards some of the least-used titles.

In Mr. Baker's grim vision, when the paper journals are mostly gone, JSTOR will hold a monopoly on their scholarly information. "No matter whether it's a nonprofit or a for-profit," he says, "anytime you have one place being the sole source of a very important stream of information, you're going to have problems with the price down the line.

"I can imagine a time when the subscription prices for JSTOR are substantially higher than they are now, and libraries will look at paying these things until the end of time."

Librarians have their own opinions about Mr. Baker. Lynn Scott Cochrane, at Denison, says he's out of touch.

"Nicholson Baker learned some years ago that there was a bright future for him in being a Cassandra," she says. "I think most people trust librarians not to make stupid decisions." Paper copies of journals are little-used, she adds; most students prefer to go online.

Kevin Guthrie chimes in: "I think Nicholson Baker doesn't really understand what we're doing," he says. Monopoly or no, JSTOR's president predicts that its prices will go down over time, not up. The costs of maintaining the archive will be spread out as more institutions buy into JSTOR, he says, noting that the company recently gave rebates to institutions that had bought in early, when costs were higher.

A monopoly is out of the question, because research libraries probably won't discard their print journals, Mr. Guthrie says. JSTOR is working informally with the Center for Research Libraries, a Chicago-based consortium of college and university libraries, to collect a full run of each of the digitized journals and pass along copies of some paper issues to research libraries that need them.

Smaller libraries, which have the most need for space-saving technologies, often don't have complete runs of journals, anyway. "I would not regard it as a terrible decision for a small college that has an incomplete run of a small journal to discard it," Mr. Guthrie says.

In fact, he hopes that JSTOR can eventually serve as a clearinghouse for old journals, collecting and redistributing them to create full runs at depositories across the country. However, work on that plan hasn't started yet, he says.

"We don't have the resources to keep everything," Mr. Guthrie says. But where there are "reasonable, well-thought-out plans for protection and replacement," digitizing is an "appropriate" solution, he says.

"And that can make the resources available to expand access that's never been available before."

FROM PAGES TO PIXELS Getting a journal digitized is more work than simply opening a book on a scanner and pressing a button. JSTOR's librarians first assemble a complete run of a journal from publishers or libraries that are willing to donate, lend, or sell their copies, and from dealers who specialize in runs of old journals. JSTOR replaces incomplete or fragile volumes and documents the contents of the entire run. Then JSTOR sends the journals to Barbados or India, where it has made deals with scanning companies. There, technicians remove the bindings of the journals. The journal articles are scanned twice. One creates an image of the page, which the user sees. The other uses character-recognition software to create a searchable, text-only file. Photographs, charts, and illustrations are scanned and posted separately for the best image resolution. As the journals are returned -- in paper and on CD-ROM -- JSTOR's employees check to make sure that all of the contents are there and are in order. Then they transfer the digitized journals to JSTOR's servers, rebind the paper versions, and keep them or send them back whence they came. SOURCE: Chronicle reporting

Easy-to-print version

E-mail this article