Stanford University Home

Stanford News Archive

Stanford Report, October 15, 2003

Digitally recording GATT for posterity Major university archive project focuses on WTO’s predecessor


In the spring of 1999, a Stanford professor and two university librarians walked into the headquarters of the World Trade Organization (WTO) in Geneva, Switzerland, to propose an unusual collaboration.

Librarians Anthony Angiletta, left, and Chuck Eckman helped to spearhead the university’s multi-year effort to digitize millions of documents from the General Agreement on Tariffs and Trade, the World Trade Organization’s forerunner that negotiated tariffs on global commerce for 47 years. “This project is unprecedented,” Angiletta said. Photo: L.A. Cicero

Judith Goldstein, Chuck Eckman and Anthony M. Angiletta requested access to the archive of the General Agreement on Tariffs and Trade, the WTO's predecessor that negotiated tariffs on world commerce from 1947 to 1994. Backed by University Librarian Michael Keller, the trio proposed sending a team from Stanford to record millions of pages of GATT documents, archival material and publications moldering away in a hodgepodge of boxes stored throughout the building. In return, the WTO would receive a set of digitized searchable historical files. And scholars, subject to approval from the WTO, would have access through Stanford to a gold mine of material on the rules governing trade between nations during the latter half of the 20th century.

"We decided, a bit on the spot, to propose they contract with us to preserve their history," said Goldstein, a professor of political science whose research initially catalyzed the project. "Far less academic work has been done on this international organization than any of the other post-World War II agencies. The importance of international trade and the rise of anti-global sentiment have reinforced scholars' interest in the organization."

Despite the sweeping overture, Goldstein and her colleagues didn't expect to be taken seriously.

A committee of the United Nations Conference on Trade and Employment meets in Havana’s El Capitolio in 1948. Historic photographs, like this one, will be made available to the public through the university’s GATT digital archive. GATT archive photo

In the spring of 1999, without a director-general in office, the WTO was in flux. The Stanford team found support from a deputy director-general, Warren Lavorel, an American who happened to be a Stanford alumnus. "He was very sympathetic," Goldstein said. In May, the university signed a contract, renewable on an annual basis, with the WTO's secretariat to record and digitize the GATT's archives. The agreement would allow the Stanford team to work on site for six weeks while the WTO was on summer recess.

From the start, the project's organizers treaded carefully in Geneva. They recruited participants through word of mouth and avoided publicity because the WTO's secretariat was answerable to its members. "All we needed was one nation to bring us to a general counsel meeting -- every nation has a veto vote -- and that would be the end of it," Goldstein said. At first, the team was treated with suspicion. The French delegation lodged a protest, arguing that French universities should get access. "The director-general said they could certainly have access if they had a couple of million dollars and wanted to come in to do it themselves," Goldstein said. "They stopped protesting." Eckman, international documents librarian and head of the library's Social Science Resource Group (SSRG), explained that Stanford, as a private institution, offered independence from governmental oversight. "Probably, if we had been a public university in the States it might have endangered our chances for success," he said.

During the next four summers, the team returned and set up an assembly line of scanners and computers in the Grand Salle, the WTO's largest meeting hall. "A couple of years in, we became part of the fabric," Goldstein said. "We went from being foreigners to being, in some ways, part and parcel of their project."

Project genesis

In the mid-1990s, when Goldstein began conducting research on international trade, she never expected she would end up heading a multiyear digitalization venture in Switzerland. She just wanted to understand how a country joining an international organization affects politics at home and decided to test her hunches on the WTO. The professor searched the university's holdings -- it had been an official GATT depository since 1952 -- and discovered "the stuff available was really milquetoast."

In 1995, Goldstein learned about a confidential U.S. government fiche collection. To be eligible to purchase it, the professor had to be named to a trade delegation, and this is where another Stanford connection came in handy. "I had a student whose father was the ranking Democrat on the trade subcommittee of the Ways and Means Committee," Goldstein said. She wrote to Rep. Robert Matsui, D-Calif., who agreed to support her request for access. In 1996, the university purchased the confidential 300,000-page microfiche after agreeing to keep the material in a secure location and give only Goldstein access. Despite her initial excitement, however, the acquisition did not turn out to be the mother lode. "We looked through it and, lo and behold, it still didn't have the stuff I wanted," she said. "It contained only very general formal documents. What I wanted was informal policy positions." These might include, for example, bilateral negotiations between Greece and the United States on reducing leather tariffs.

Stymied in her quest for answers, Goldstein in 1999 asked Angiletta, the SSRG's social science curator, to visit the WTO. After speaking with many bureaucrats, Angiletta talked to an assistant who had worked under three GATT chiefs: "My first question was, 'We understand there are these bilateral negotiations. Is there a written record of them?'" When the person replied yes, Angiletta politely asked if he could look at them. "Oh," the assistant said, "they're across the hall." Angiletta was stunned when he was taken to a small room lined with carefully bound volumes of the entire history of bilateral negotiations going back to the GATT's first trade round in Tokyo in 1947. Finally, the object of Goldstein's long search had been discovered. Angiletta was also shown a room crammed with rows and rows of old documents filed in boxes that had never been recorded on microfiche.

"He came back and said, 'You can't believe what's there. You have to look at this,'" Goldstein recalled. Two months later, the professor and the librarians traveled to Geneva, and the GATT project was born.  

How it worked

At the beginning, the project demanded more chutzpah than know-how. "The first year was total hubris," Goldstein said. In 1999, Stanford had experience with boutique digitization -- it could handle a book a month -- but no large-scale projects. The group learned through trial and error: In 1999, they scanned 150,000 pages; by 2002, the figure jumped to more than 750,000. Each summer, library staff would pack everything needed for a remote scanning project -- from computers to duct tape. In 2002, the shipment weighed 2,358 pounds. "Because of the diplomatic and political conditions surrounding the origins of the project, we had to go prepared for anything," said Stuart Snydman, manager of the university's digital library project.

On site in the Grand Salle, staff set up an assembly line of 11 workstations, two servers, nine scanners and two metadata stations for cataloging material. "It's a whole production," Snydman said. "You go from metadata entry, to disassembly, to scanning, to quality control, to reassembly, to return." Unlike legal documents that can be scanned automatically, GATT material had to be prepared by hand. It was painstaking work, much of it done by students. Each summer, Stanford would hire from 25 to 75 temporary staff from Stanford and other American universities to work two-week stints from 8 a.m. to 4:30 p.m. or from 4:30 to 11 p.m. Students had to follow a dress code -- no jeans, T-shirts or running shoes -- and they were not permitted to wander around the building.

The Stanford team plans to return to Geneva next summer to scan the rest of the available material. Key documents still missing include the personal papers of the first three directors and what Angiletta refers to as "the mythic one million pages of agricultural material." Pending fundraising, the group also wants to conduct oral interviews with about 25 GATT employees, many of whom are retired.

Project accomplishments

So far, 80 percent of about 2 million GATT documents and hundreds of historic photographs have been scanned at a cost of about $750,000. "This project is unprecedented," Angiletta said. "No one has ever done a digital archive project of this dimension, to our knowledge, anywhere."

The project's scope required the university to invent a new, high-volume, high-quality production-scanning lab able to record and catalog specialized, one-of-a-kind documents in a remote location. Regarded as a touchstone for library digitization, Stanford has long-term plans to use the experience to develop a "service bureau model" that will enable the university's library to contract with other organizations seeking to preserve their history. Snydman said the work puts the university's expertise in a class by itself. "There is probably no organization out there that can combine high-volume production digitization work flow with the specialization and uniqueness of creating an academic archive," he said.

Stanford has played a critical role in preserving the GATT's history, project leaders said. Operating on a slim budget, the GATT did not maintain an archive or routinely declassify documents -- ranging from everyday internal correspondence to highly sensitive informal tariff discussions. "One of the amazing things about the GATT and the WTO, both in appearance and in practice, is that they have not been archetypes of transparency," Angiletta said. "Many documents that might otherwise have been public ... were restricted."

However, since the riots associated with the 1999 trade round in Seattle, WTO officials have understood the political salience of promoting openness, Goldstein said. "[Officials] are declassifying the new WTO stuff much faster, and they are taking up the issue of declassification of the older material," she said.

In a move intended to bolster public access, the federal Institute for Museum and Library Services in 2002 awarded Stanford a two-year grant to digitize the unique microfiche collection and provide general web access to public GATT material dating from 1947 to 1986. According to Eckman, the project is scheduled to be completed by February 2004 and will give users remote access. "We'd like to have a way for people to get permission to use the confidential material," he added. "The WTO makes the determination, although we're hoping to support transparency."

Without Stanford's involvement, Goldstein is convinced the GATT's history would have been lost. "Old documents get moldy and eventually get ruined," she said. Taking such a proactive step is unusual for a university, the professor added: "We don't usually create knowledge in this way. Normally, we have a very passive role in thinking about the materials we study. Here, we were actually creating the materials."

Eventually, the GATT Digital Archive will enable scholars to tap into a wealth of information on how international trade developed after World War II. For Goldstein, her deepened understanding of the GATT reinforced earlier hunches. "There's been a lot of confirmation of how I think the world really was," she said. "Trade is political from the top to the bottom. It's more about politics than it is about economics."


Judith Goldstein

Stuart Snydman