Google Scans, Contract Released

Search firm also will link to archives, allow downloads

By Andrew Albanese -- Library Journal, 10/1/2006

University of Michigan (UM) officials have offered a first look at a program called MBooks, which uses scans made by Google in its partnership with UM to give library patrons the capacity to access text directly from the university’s catalog. Under the program, the online catalog will point to an “MBooks” button that, using Google’s scans, allows patrons—in the United States for now—the option of keyword searching within a volume and retrieving the number of times a search term appears per page.

Users can then virtually “flip through” the selected work using a page-turning function. It also has enhanced features such as zooming in or out on pages and rotating images. Although just a small number of UM’s holdings are now available in MBooks, co-interim university librarian John Price Wilkin said the library is moving ahead at an “unprecedented” rate and eventually all “uncopyrightable works,” including government documents and works in the public domain, will be available for full-text viewing.

Works under copyright, or even assumed to be under copyright, however, are not viewable. Still, UM officials say, the capacity to keyword search within copyrighted works represents progress.

Meanwhile, the University of California (UC) reignited librarians’ concerns over Google’s scan plan after it released the terms of its recent partnership deal with Google. The six-year contract will involve at least 2.5 million books, with UC providing 600 books per day for the first two months and more once the project is up to full capacity. In return for its participation, UC will receive one copy of the scans. Critics, however, note that the university can’t share, license, or sell its scans to any third party and can redistribute no more than ten percent of scanned material to other libraries or schools, even for educational purposes—which constrains interlibrary loan.

Internet Archive and Open Content Alliance (OCA) founder Brewster Kahle noted that the contract shows there has been no evolution in Google’s practices. He stressed that the Google project is not a public resource but “the private library” of a single corporation, while the OCA, which includes Microsoft, Yahoo, and major university library and publishing partners (including UC), is committed to openness.

Google also announced a program to crawl content in magazine, newspaper, and aggregator databases and return results in Google search queries. Those results will include both free and pay per view articles—mostly, it appears, the latter, though results are not sorted by price. Since most newspapers and magazines have already digitized their archives and offer access to their back files both directly via subscriptions and aggregators, the plan could boost archive revenues for periodical publishers. Initial partners include Time, the New York Times, and Washington Post and aggregators Factiva, LexisNexis, Thomson Gale, and HighBeam Research, among others.

Of course, as Gary Price noted on Resource Shelf, much of the content patrons will see listed as pay per view can be accessed for free via libraries, which have already purchased access through license agreements. The onus, however, will be on libraries to let users know they can access content returned by Google.

Google Book Search also unveiled two innovations: one provides links to WorldCat.org and international library catalogs, while the other allows users to download out-of-copyright books. So far reviews from librarians have been mixed. Downloadable access seems to be hit or miss, with no real pattern for finding which books are available. In addition, librarians noted, the quality so far is average. Users can expect improvement, however, as the rollout continues.