Yale online book project hits setback

Julie Brown

Keith Wight, an image technician at Kirtas Technologies, monitors a machine that photographs the pages of books. Kirtas is scanning books from the Yale library, but project sponsor Microsoft recently said it is abandoning the project. View full image

Last fall, the Yale University Library belatedly joined the trend toward mass digitization of its collections, signing on with another latecomer, Microsoft Corporation, to scan 100,000 books over the next year. Microsoft was to foot the bill in exchange for the right to make the books available on a Microsoft website. But in May, Microsoft abruptly shut down its project, leaving Yale and the company's other library partners unsure what they'll do next.

"It was a bolt from the blue," says University Librarian Alice Prochaska about the decision. The company said simply that it was planning to focus on more lucrative areas. But it was widely speculated that Microsoft decided it could not catch up with Google, which had begun a more aggressive digitization project a year earlier.

Although Microsoft has now dropped the online book search tool that until recently could be found on its website, it may continue to pay for scanning of Yale's books for another few months. Prochaska hopes that 50,000 books will have been scanned by that time, many of them volumes that exist in no other libraries. Yale will make them available on its own website and on the Internet Archive, a nonprofit repository of digitized books and other media.

Yale's books are being scanned by a vendor called Kirtas Technologies, at a facility in Wallingford, Connecticut. They are placed in the tender care of robotic cameras that use puffs of air—"gentler than the human hand," Kirtas promises—to turn each page and take a digital photo.

"We're disappointed not to be continuing to work with Microsoft," says Prochaska, "but we felt they gave us a good deal, and they gave us what has turned out to be a free startup for a very important process that we do intend to continue." Prochaska says that it's too early to know just how the university will proceed with mass digitization.

Besides the obvious benefit of faster and more far-flung access, vast virtual libraries might also open new kinds of scholarly inquiry. For example, Prochaska says, imagine if you could search the full text of Benjamin Franklin's journals by keyword and place, search the full text of other diarists of his time, search the full text of all the political theater and playbills in France, and then pin it all on a map of Paris. Seeing who was crossing paths when a given play was staged might deepen our understanding of the political discourse circulating in that period.

Data mining like that, Prochaska says, could mean "whole new disciplines springing into existence, using computers to process data so massive it was impossible to handle before. This is pioneering stuff."